model_id,model_name,provider,weights_status,selection_rank,benchmark_id,benchmark_name,benchmark_category,measurement_id,run_fingerprint,protocol_fingerprint,benchmark_release_id,benchmark_version,split_or_window,metric,unit,higher_is_better,score,score_display,display_role,comparison_lane,reasoning_profile_id,reasoning_label,reasoning_effort,reasoning_enabled,reasoning_strategy,reasoning_token_budget,reasoning_is_provider_default,freshness_at,freshness_status,age_days,trust_rank,trust_tier,source_authority_rank,source_authority_tier,evidence_rank,evidence_grade,verification_status,date_rank,verified_status,source_id,source_name,source_url
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,epoch-eci,Epoch Capabilities Index,capability-index,25820f8dd2ecd9ae44b6417ccf946dcfb5403ae772fb9cd77d496eba71baaf5f,d887192a563872c70d5ac00a9814016437b1d4d5f425658ead9fb49ae8c7660b,7ebfea7bab2e4747e9bb4b134bc1e2a358408c8e41c568da62257ecf353d821b,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,169.06,169.06,exact,canonical-source,profile_9276dc96b9e1782b83ae6c86,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,bd3f4b9127522f85eef17335009ed080803bd901e8c4227e329e88eb87380562,b2d83c91b700f8484fd63b3d9703405d9d5a1235f8084606919aa286c43c9df8,ff3aadf4e201cbeb10a8718d8c05c6c1a72ddb0a0cfbfc86f35e5ae115c2acac,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,56.853856,56.9,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,arena-text,Arena Text,preference,246cfc9f8dd4aa4f572c730f0320030513e1993007620ffacbcd392ea332c15d,f3fe335d198407b8eae129b636da4d7cb87c20cb921cb9b8fd4ebfc7d4bdeca0,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1475.647885100779,1476,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,livebench,LiveBench,general,413c2ef9e8360cb6fd8bdc28414a39651a7bf9cca804d565df431b3935da031e,f6ba353da222a95140a834495556445c7241a83d1f3b73789713cca425ca51ba,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,82.16064285714288,82.16,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-09-05T16:42:11Z,fresh,56,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,ed46ba9f7f7814d1dfffee4d17b91d7b62fa2b6db556d471d2051de962192408,e21e3a7a3cec15d944adaa2566a20978cb26beb9fd50e624842a014c5c65f4a4,425e0d0c3e4fa27c86e206b700698ff0a141e9c7561ce80edae61b0fffdd0616,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,54.8,54.80,exact,canonical-source,profile_38556cdaf0f3888ee868b8cb,reported,unknown,,unknown,,false,2026-09-09T18:59:21Z,fresh,52,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,5b76c510cf20f46f89f480f44245a40f716e1d3a67a0c4e2cdbc908e596e2af7,2e49cc198a4a2284cb72277bd6dbb9a5673b863a94b3458984b2c520e3f91bf9,2af561e3547a161f266c858e2da266c8a3686a586d0456e21e5863b3b54bc181,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,54.680259,54.7,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,simpleqa-verified,SimpleQA Verified,factuality,ba487a803e8367590119e1033cb332fcb1e32fc9ad9f27881da40ad61c570e91,df1a193413ea3fa2fa0794e0369663191350e47f41e35a586efd953f6374b626,2854e3d43358f88d6e47f693d81162b933415b351afa2cdfc215f08a78adcd61,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,75.6,75.6,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-08-30T14:57:43.000Z,fresh,62,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,gpqa-diamond,GPQA Diamond,science,4a0c0baf62127c909ccbb6a4734d17990384e9b8471a2244694d088e4f168735,d207c38bcf93ca0233d9a7c109c14c1a124df2c72f3ebf981768d5cd5ca20c1e,59087f63ed4c023451438f8ee8ba33c99013eaea3114b59436e916996ebbfd44,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,96.060606,96.1,other-supplemental,supplemental-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,c133a59aaea1182115fb6cc7bddd8c4d90b567eef3db965b41a1cc71334d9a4b,0d76f06dfb6b9d946c231f3a49d761768417f3deb68d7ce20a75c4a1ef2f54fb,966adef4976badfac20721ce0a1dd126f88966c8309933d9a2fd20e6aff1f86b,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,100.0,100.0,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-08-30T14:57:43.000Z,fresh,62,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,frontiermath,FrontierMath T1–3,mathematics,40392a2fabb1edd72abef4d132a67bcb41a3f5b6a36cedec884a66b8a1fa5cc4,6fd8f32a8a063a7ff44cfb20d6b6aef57a952cb5dd5a3601c30e7ebd66962130,313848536ae7469ee53b15def714d6dcc5afc20f13f3509167c3e5b0871cedf3,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,93.6842105263158,93.7,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-08-30T14:57:43.000Z,fresh,62,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,arc-agi-2,ARC-AGI-2,abstraction,22416d4e9d88abba95b2d4906e3d146784c67760348c189f584e11f6c3645db9,9b7dc98d98bedae0a63458bc64a7ce49fe29bc6500ab28488bc5208ca70ba9e7,b28d063f4e06ef5ecce16f5af8d1f6ae972c1cfcc5d40ff95a612ee586729b39,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,95.0,95,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-09-03T22:49:18Z,fresh,58,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,frontiermath-tier-4,FrontierMath T4,mathematics,7865a44e9658471fb16e762c68809b9e338f2ccdaece3cb91f7edadbeb35d327,e1ef89374da6b91c2d8faacb07535630fe76623009011266307d5783b3aba45e,04a3a0352ec5ad74f3c30888cdf8e8bb18489aa80e36e7b0c0f77af63d2842d8,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,97.6,97.6,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-08-30T02:22:12.000Z,fresh,63,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,mmmu-pro,MMMU-Pro,multimodal,c4f777d86318d2e0b37f270728e9aefdd8da6010189e6279370b3d69279bd69c,99ec5413a659f95420f1708928d795734597c527bd93a9cef0459e4c48eb77e6,5a7867656e43542c6cc90fd79af05ff5ba7a2136e56a3fe40ef8e7878437f909,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,86.878613,86.9,other-supplemental,supplemental-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,critpt,CritPt,science,41acdc62087a6a65d16751f288cd532ebb680fd37e425e80982e23241a9d9b21,46019d866e6bcfa2a7365749a933c5f0c88daec3682cf20d04a6e6151f1d59e0,3eb901ca4f835e5f8e567b934a3a5f206c51f2cee95c93b16e1a5cf06633e2d5,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,31.7142857142857,31.7,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,chess-puzzles,Chess Puzzles,games,51f0da173a8628fcb7699a66cd0e7f0c739fe39c3dfbe76a181fac2684d4003f,24363a4a21b6d7180bfb4c8ecb60e23a22a2b0bdb3bcc8519792e76a3434f0ae,86f9ec27db74333671b2d764f3712aded33e58fbf4c4bcd07cb6a6f2413ca95c,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,74.0,74.0,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-13T13:42:38.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,arc-agi-1,ARC-AGI-1,abstraction,87725e7b6ad28920efedc3e609084a3b882a9cd4e0d2834beeae56e488cc00e5,ce6574aa279e06198bccec76c7f78a20b82790d4201380548821c2469451b5cf,d34c2d359536016f27e3aff27e6b7bc679d90fe39a26d13dbeb96ff4547f5eeb,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,97.5,97.5,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-09-06T19:02:22Z,fresh,55,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,deepswe,DeepSWE,coding-agent,52da400c4513cb966dad2116a772e4a6a055a9abc03486fa0f2102253ae1ae18,3b9fde1893a272dc3b59dce4eb0ab69a9fda40d39d87d5fcaed61eae87bde541,d4e26041d9b76d03a07c7366187d69f1095660ca79d9891f8caf0c00b728abb8,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,73.23008849557522,73.2,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-09-04T01:03:12Z,fresh,58,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,89ef1d67cf2c84a6e77a182ac34f784fd874f243bd0f4d975c10d36d384ea318,4cac5cd7de8a00a2c6f15b3c80bddcfb1c635e4765c259240d1f660b4d602230,c5b7edf07c0dfeaf1ea2f623c2b4c5092011aaf405beab9a4cc3640de33d0f61,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,96.9,96.9,exact,canonical-source,profile_a8eb1aa90528f25d7a199e22,high,high,true,fixed_effort,,false,2026-10-07T01:44:58Z,fresh,25,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,gso,GSO,coding-agent,be8f99a18b01cc90b97b3d83fe86a5194b5d1ac41c83c299c964b332aada9e22,46dd1c7ab248af99bce0dca5d42879c4d5dc59671492329fd9e596404bc303d2,eebc9651e1990fb52f6f206b1084d65a0640d7c2d9774a2676ec8676ee434518,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,79.41,79.4,exact,canonical-source,profile_373c486c9fe4fc91578ca390,reported,unknown,,unknown,,false,2026-09-27,fresh,35,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,frontiercode,FrontierCode 1.1 · Main,coding-agent,4b566fbcb87258b345d372b5fe07e9dba84004e009b93dcdb4385adc7a36a7f6,2d82ee0270e995af682c4778552573bf76e9a6661b668da79f4b2b1f70208f98,e19b193e393c9e81b783925a453da14203f8cff7a54b1f5af520cac4f7770b2a,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,53.26,53.3,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-09-06T19:02:21Z,fresh,55,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,ioi,IOI,coding,bbc37cb201b1665196fd0ae262f4ed8e8054f6a8e5637c03c223b87d857cdfb0,2489de89e66c5446f66d11127e12349674f7de2fcff9622d4a9d23ad6b9dbcda,7a6eb621851bec0b74e66fba6ba514b3183dff39a19a47ed8fa5e55b05b86f8d,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,100.0,100.0,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-09-14T23:34:42Z,fresh,47,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,scicode,SciCode,coding,f57b923423daeb47cbc0735cb5c5cd39642628e17dea035f74220e5f58e119ae,d114cb371ebb440f250faa69648bc7bade19bee1ee6b1522621114aeec4d2eda,ee39d1995ea289ec8c14a15e390ea330a1f94e4f235f923ce2d7a3b0c8e6b5bf,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,56.4814814814815,56.5,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,ale-bench,ALE-Bench,coding,0d61f2caeeb99a1fc0234c9d43fc69decbe2688a742daed736cd57fb2843f227,319b24c34162e26cf1539535711df172934ad236e5995f1dfb6d420dde739ce4,c2edd8d4c38660c681e4fd4857862e563cc9344321c9f62a8287a3235d44a90c,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,2951.32,2951.32,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-09-30T18:53:16Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,vibe-code-bench,Vibe Code Bench,coding-agent,f69fadc2dad38e91d06cfe38c8fc3ebb33680899c0345cdcb5243faacb6ab4fb,3168256c1f6b7de0ac4c3bed943edd57a1a08c31ce2058985f706ada1c18c011,41892bda746320939fc120ca2546e4907455517f55ab8144325f2fce073d3f41,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,89.594,89.6,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-09-04T09:09:25Z,fresh,57,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,webdev-arena,Text Arena · Coding,coding,11ea0b38a8e9c9783f093e9aaebbf9bcc6156fdbbf5b1240ff13e1b26eb20834,67984a98170624fc9861869b648376752c130fc521bbe5f27abc896660ef73d6,442422ad0dde6f9afac723800b517a077c6cd85550c4514675179008aa01f345,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1773.3545606290113,1773,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,mirrorcode,MirrorCode,coding-agent,a63dabde48a505aa7a36467f90919df747a8eb3d251abf8c2a866d09a62b6121,f3c0ba72c6a41dec217946658746864f5d0c68c6852f2ed71a38de70157979c1,712acdabde9fec9e9674eeab8ceacc45f20e87c6f37ca1761015d0e8eb290bcb,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,60.0,60.0,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-29T18:04:06.000Z,fresh,2,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,programbench,ProgramBench,coding-agent,cfb49952b20b10ec546cac567c8232779b36bc6f92cab9e07f650a0aae175ed0,986773133f9bfa0747fdcf125b5408fad239a71c754b0a64d6bea6d334629378,95554c804b3d3488cce1c61c1addef0b15f5eb4e2cb658476d8c8a4a69945000,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,5.5,5.5,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,weirdml-v3,WeirdML v3,coding-agent,63e379b8ff7277a43793313fdd70b0505de49c63d054dd176b414e0e082077de,e9a693818f0ba08a7f2226b81c67d0b6c8c227717e384a279b836bbfd95dfdef,74b30cd6fbe5538aa72458734b83df5bb9ca837e739ce9b9600fbc373388e1e7,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,42.220231363636366,42.2,exact,canonical-source,profile_68496eb7a3dbdd992be334df,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,8faf7df5250b61b0138c02e34704e6ff9ee6085d5efc172b5bed6af6f6386b0a,31a5004faf4c16afc592f51a1134f701526d30cd619a224114baa69a0b405e3a,b2cfc1f6453633f55ed1f444420221deb0c3736a0f4b94755fdf388c115c2981,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,87.266,87.3,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-09-04T09:09:19Z,fresh,57,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,49129a3f85dcb70cbd083dbf5e3c90bda29e12ea1d3aaddac9f9e2b107c9a99f,031909787bf8cbfbab478b69b4e8748a290c1c262b57100d5c09632c337717b2,d7d11042790079edad82bb1abd1aad8600cdc1f8e3b286cfe771aa8bd20d38ee,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,59.596,59.6,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,sre-bench,SRE Bench,terminal-agent,5f3381dee4001c3a4be7550a39f91b25c173852e58ffc26b233a7f99529187be,61834725e3c31bee87b533ec57ac834fcb1c618b504b00ec964f6d0d67ea78e8,d237b1e83538dd1fc10db7618f65ecf6adffc3c8b022e5b4713fda66bf16d985,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,56.87,56.9,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,f5eb9e107c3be4d5e8f75c89d74e1231e591ededca82c0fcf8796cb6301eb491,704e8dddd83992489aaad8301110dcbf7e4403f5092ec9c96d4b2d542f61eb64,f9840e8c650299f0b3ff78d44f36cf5ebd2dfcf465799587b93c39759d11f106,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,41.443299,41.4,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,apex-agents,APEX-Agents,tool-use,5c458526aef24450b7dc29bafa5f2c4a83afc1e18a394140450e76650a46a8de,f1ee9fddf4756444a826bdaaa3ac12346b27d9a61091ecb2f803d5851da248a3,925e7392b66354d25392570cb0544fe4125cf86bcfee1eb0ca1ac3d5cfab2995,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,64.7,64.7,exact,canonical-source,profile_138452291adbe4025c01d25b,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,arena-agent,LMArena Agent Arena,multi-turn-agent,73c1ae1be1eafe3f5e1163d8b16f71865c6327b57d0b16398034c019075e33ba,7a11c18cffbd75c479f344df49aac1bb85d9e833bfa80c437a59547f4496525b,db0617a03c49a34a923e2f8ffda6cd267cd5dc66edc46d706ff106876c9d1ff7,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.12960129693659303,0.1296,exact,source-specific,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,vending-bench-2,Vending-Bench 2,multi-turn-agent,a7ec89e84db94833c2358299cdd8fca8f682cb01823622b2569276e3f2bdd0a0,b5ab987f1d03f3511c4ea27caee959af6eaa636b8ea5c9eea6b5a3eb83781b2e,cc9fb18a90926b80745487407023b43b6b38c6beb0a76f23d2522810152ae557,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,15514.701666666668,15514.70,exact,canonical-source,profile_2315d6296613da07b43f9c55,reported,unknown,,unknown,,false,2026-09-30T18:53:16Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,36e9084a51a24b9ca343519fb5f061cc0701764160cbcec5593c612eb5f03c4e,f8bb22f60f8544e280e74ffd6bcb69d2e7c6041669ef2d7b16c557a7adab66b9,2e91e2e91472eda21f4f7e0f089a1651686238263436658a4d19e3c86ac43aa3,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1583.69,1584,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,17bdd8840d010d19aa2b4dc26d0c1b36682c3b6f1c692e86eee7a3b37b024e85,7a12a4379083ae05c430487f76bd7b0fd535c281bec50017f99b59e23207a6bc,d05ec267b780ae1aaa7f08e1be2af16055c12d7d811ea626692c9f839a2aba19,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,62.857,62.9,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
openai/gpt-6-astra,GPT 6 Astra,OpenAI,closed,1,weirdml,WeirdML,coding,231e040e4238f44ea49e19e998d2ccf85982c9b99c8880fdd4336ea3a3adbe83,90d8f867f78de4c4c162946486b2de38c513f00112804f23741a5fa93f7d200e,8ef8215c0ebf37651b14f2e485100d791f62dd592f8b348bbe2f92dd1dc5c1cd,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,93.57,93.6,exact,canonical-source,profile_a45ce5352161ba5d3a87a651,max,max,true,fixed_effort,,false,2026-09-30T18:53:16Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,epoch-eci,Epoch Capabilities Index,capability-index,b4e06c205ff0b0f90dada7948d8da6f7559199fa8f63237e3bd746c1e860ddac,55d340cdebe1ad21837614a2d6585185c880528f263d9064773d75deefd61233,f3fa80a685806f77c42d904ab340df7bdbd1096fdb1f68a5910bae883b09246a,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,168.4,168.40,exact,canonical-source,profile_f9bccce8d4fa3cdf550e517a,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,25338822491a5686002b0607491a6ca8feb1dc204140293c35a75ea05dcd7cba,a8a5760096d595b49eeb49e787e1dd8df989d8a35cc70f3ad081fd634e7f75fa,81ebbedac6bb686cde756f69be10b88f093baa09cc773496800bebf083fb3cd0,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,58.406034,58.4,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,arena-text,Arena Text,preference,2274a8e4d22f18b621e653a8fac635a716c008265dc9d0f19e04b01f3af08304,2c5ccaf9ad09a70a31c5b4a49cffc25c7d97071c6384f42fc3b3d3b4793afbdd,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1508.697590625909,1509,exact,canonical-source,profile_37e371ca4768fbe805410768,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,livebench,LiveBench,general,b64ea79f2b88710103459ae2364e6e01a3a6a285ea34be265205e89e6790efdb,cdf5d8b03bc09f2ec3259baa3a2141bb4b8936f73c3f78261407c8b8177538f6,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,83.21657142857143,83.22,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-30T18:53:19Z,fresh,31,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,a158b9c3ae8fb3ca5325fbd4ba8eeca36a3bf0c5624e3cd29a5af4863e1f97b3,dea35a53e7d991c4c7c4b3fc71f8155cb2a035ecc7eac600518c0ce68ded66eb,4bb35cdffa7da4653c3f76ea0c475ae842608d588f64ddc5a5e4a0f9d1bde2c0,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,61.353105,61.4,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-08T22:23:24Z,fresh,23,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,simpleqa-verified,SimpleQA Verified,factuality,c11471bbb41780b90e6097d94fa891e1863026ba4991a704d8c11db973805250,a9c3f5f2f38cac847d801519da2e4f6b461a954c4f5e352efaa7df692a2fe384,63ee90334824e9cf59ed152b9627654b2fcb965eef7037ccf787c24b853603fe,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,72.2,72.2,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-22T18:27:39.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,gpqa-diamond,GPQA Diamond,science,506c1ad91eefca4538eaa020cffc94da8f4b29430616d804980370df30776069,37c40d2f488d30bdd4bbe22f8b5ed20ac85bfa7aaeeeac947817052fe05b5470,919ebf8a3d132d097dbb0cc21c7ba269afb98b98144deb645c0be6e6390220de,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,90.59343434343434,90.6,other-supplemental,supplemental-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-22T18:27:39.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,f1d81ce5b474147f6223c0914ee1da9d202433bef69daf8ec0d454e51d3e19d1,af451d1fe3928df696e53118c61ff3c5ce5feeb0d4335805774302963a1bc7b1,4ab8432dffdd095ab10181addaf3fcc8c8f586ceccecd8dbf74f176305c5c768,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,100.0,100.0,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-22T18:27:39.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,frontiermath,FrontierMath T1–3,mathematics,62d3803fc8f10c86a7c1f7858040cbad0d50e2330a464b9beadc41ebf34fa0a8,df2796b9b6eda77e8c52a24b33f141535ddcf83fe83077f4687398f818541780,4e2bced60f1ef278854d63ac4e8ca31823c2c92c09c92122c32905e6348ade6f,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,91.22807017543859,91.2,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-22T18:27:39.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,arc-agi-2,ARC-AGI-2,abstraction,1eeb20ed84f86d957135d77863cea385e5fe92dc4d3fca9ecc45bcdd13515bbd,96309dd627582d04cc724cb621097e4f8256a417e6fccaf5d7d84c49397b51d7,315b4f3d86330f1b0ad624dbc9cb71b7d38a32a3451c9f98622199f6464ab2d3,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,91.66666666666666,92,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-30T18:53:20Z,fresh,31,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,frontiermath-tier-4,FrontierMath T4,mathematics,1a9ba1d68ce3da33282f8592cf0278f68b87a369d69331ad35b2843f4f1b3d96,0e41c38fa862ca88746933f8dba99c99a1d153f1338f2c7ec44315f08545c9ae,f0f19eb269aa45592a3427a4bb852adbd365864a3ad88b0c430a42d42ef2d8f8,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,95.0,95.0,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-22T18:27:39.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,mmmu-pro,MMMU-Pro,multimodal,72c0c7154e9395931ff4bc2fca733903c9aeed35656c48faeb983947803bf0d7,0bb9c84e99fa02f1d32cb57c8b4a79df9826d079150715c6cd1d688bf6826ecc,c460578f5b259e5a325bcf140810789b6667c21a1cc093827483293ad4510cc6,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,87.687861,87.7,other-supplemental,supplemental-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-08T22:23:24Z,fresh,23,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,critpt,CritPt,science,32029d217e7adc76ccd4c4ad27909f64a410c9d669a2f05834ddb00f94f299f0,e5e071ac92c352bcbd2d233ced5b5ddae009cdc2a61ed1ab11ec7844c3704fae,56fca9bd3c46f9c3222f92eac14614023fb6ad69e553a40dcdc6720d8ea97108,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,31.7142857142857,31.7,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-28T21:25:03Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,chess-puzzles,Chess Puzzles,games,d99e05caa99fcc4310088198da65aeeefb79ccefac92610d9dd2a01eb9fb285c,15f181eb321c6ced3a956777c5f76f1a15e397eb69c142d85fcfa627e69b5165,7c87a06d004a39ff6bfd5f6513e0cfc6f673a695f56e189a064a27b1bfc0dd07,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,67.0,67.0,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-13T13:01:31.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,arc-agi-1,ARC-AGI-1,abstraction,5a21c9bdd54738037e6897c97d43ddb14ae467074ad06492f71219c8ab639d80,7b4285aba63a0e7c39b0ecd7fb1a68e36e2478ea1a41ae0a067fa57697e2c33d,55c1379708f54011fe6209562f23097353ba4c1673e14d76e4b935194212a208,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,97.5,97.5,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-30T18:53:16Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,swe-bench-live,SWE-bench-Live · Lite,coding-agent,db978663dbb2e8f823709bb49eec361cd8b5342b51f92c98f9e4f39b2afd5977,9a3dadbe0b29ac5b8d40c45fd59ac8b0e3a6f22b96acb1525362a0af65952a00,ffb549af0b67a5007b617b9499e72d7ec536efe47268a7c134ee80aa49eda3ea,release_9d2de9d5be7e9822af19f7b3,SWE-bench-Live Lite,lite-300-python,resolved_rate,percent,true,71.66666666666667,71.7,exact,canonical-source,profile_ac8995612553984b32210e23,xhigh,xhigh,true,fixed_effort,,false,2026-10-21,fresh,11,1,A,1,A,0,A+,official-submission,3,official-submission,swe-bench-live,SWE-bench-Live · Lite,https://swe-bench-live.github.io/reports-0605.jsonl
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,deepswe,DeepSWE,coding-agent,5dcd2e5bff9d17d797b12928119cbbf0f7be536d18671dd3ef7e6f01d750f4c2,78cd262e0c7bb48c44fd719220c0c73597c4d613e1bf9c309f8b7f57b52b3379,a5cbf75f23ca816482032c8d34b067581919f96816decdf209092b05efa865f9,release_12e8a62d68d933c29dcdf020,DeepSWE 1.1 vendor run,vendor-model-card-run,resolved_rate,percent,true,74.2,74.2,other-supplemental,supplemental-source,profile_fb3d58069785743e36e95043,max,max,true,adaptive,,false,2026-09-22,fresh,40,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,anthropic-opus-5-5-card,Anthropic · Claude Opus 5.5 System Card,https://www-cdn.anthropic.com/fc1b44717c85dc068bc6ba5024219938094694bd/Claude%20Opus%205.5%20System%20Card.pdf
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,cursorbench,CursorBench,coding-agent,befbf255767fc01bbdad4c23aac601be6e1245245f1c4d29b3f359c678be217c,a6222044d80d0ecd305e15bae26bb18ff6284b8e24dd887da98e7ebec361f291,83d1cdf2d778ee124394c10baa2874ad50c7d194434194fe9bd11c556416228f,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,57.8,57.8,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,frontiercode,FrontierCode 1.1 · Main,coding-agent,d5627d58d93fe374d8471a2f56715984db0e29645c78f2b5b0d29b4273c90c31,b03012e398d7252eded9f4428abf41aa1a2633afe3a79e9c40f24425de96f261,f47b5fcf31c037c3984dae6d00e33dd6a78c5acf154abd1130cabe14bfdfd4c4,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,54.64,54.6,exact,canonical-source,profile_9bcad8178005dc3a63c077d5,medium,medium,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,ioi,IOI,coding,9a6e835ba7cb6b1537832a8bd4f5ffa7a601f621d67ca87032480cf3e07dc27c,9b7b952dd1d52f9597f4ce2f2ab974a28e3bd7a4420cf762ab431cc52344932e,7b7b55521975458bf69e2674951bf4ad411810c371182f987b6b761e6131283b,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,95.056,95.1,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-22,fresh,40,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,scicode,SciCode,coding,fe3d897f43bf8cffc5e5e4a6e5cafd86218c2744de763d2f0f5d5cdd8d010de3,bea384eb645df97fa8901edc56e8a3d5fefb3179b7eb53ba672341f644e59385,624755413dd473f97bfa598805569a937857b4644a94b64e0348aba02cbde007,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,66.8981481481482,66.9,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-28T21:25:02Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,ale-bench,ALE-Bench,coding,2448ce38ffa3c59e36214ba49ce3995cbe9653589a507ae60e6654c1e48c3e6a,43b4a6da6ca60774d6b36ac7158690000052ac4310aeb80122433058bb55f7f3,9482b517aa668f1b1f7e79656e624fc554b91de96483745271e9193220171c39,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,2147.38,2147.38,exact,canonical-source,profile_37e371ca4768fbe805410768,high,high,true,fixed_effort,,false,2026-10-05T10:25:48Z,fresh,26,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,vibe-code-bench,Vibe Code Bench,coding-agent,c920b94d1841091f06765bee5d5f76ff60f9fb3f04fe1b1a1dbb2adf6650c092,766ef172e0f2e578d25e17bcf704b5dba742c0194c737e58378624a2fbb893db,cba944f0ea0e1d5985d9068e9134f5529095f8e82695f6146d1ebf7bdd2fde65,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,90.294,90.3,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,webdev-arena,Text Arena · Coding,coding,511e8b5de685f77217b6cd2b86424d861b892173b2654d3b1fef45535301cccc,205440804ffe5a16efdcc2516998a8e6016f52820ba5fc43720089c1f2a00d66,943a6b04c5ed0ce14cf9cf8446203aa7065ca94fab1753dc2a57897ae87ad14a,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1809.8391605525123,1810,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,mirrorcode,MirrorCode,coding-agent,f68d9c6744db213647022de20be1ab9404a4f79c6b1da0836a432ea1b30c5331,fc1191f2a0a28111f2be51aaccbe22a4905bff5da1fd24fdaadb6f3b865cbe00,16c548b388367f0b7c76db1763b5889ec1e4d4a986b20227b16a4569ff7d348d,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,77.4,77.4,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-22T16:42:00.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,programbench,ProgramBench,coding-agent,03f1cfd3cd42733cc006eba5c6c821838facc200288fc2aa0a64f0c86e164562,3ab135ec55bcb491f277049fbdf047ba804d46ea41d3e6cf92c38cea25fc71b8,07fb234dfa73bddc1508bb05a8bf4eebd69e54ac282f5fa1003cfbe6f6ffa78d,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,18.5,18.5,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,weirdml-v3,WeirdML v3,coding-agent,dca96e0a47f9348e6b9d2c99a5e85beaebf86635a287f2acdad55e641a6b7776,da671ad38a38519cddf2d1cd0d5c96ed3c98b906ee582c2da6af6912216c1dc1,ddbd7e67ba37d1654b6878e6bd07510202c2db31a4fd07c7af992fe723a0528d,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,31.20028142857143,31.2,exact,canonical-source,profile_ac8995612553984b32210e23,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,7c76a3394c9cc587309591a9d38eddf5a610e9bb80583a668f997edae39101e3,4aa566bf935ccb8ace357734aaba310af2b2afc731d32946ea575b837cf4ceb8,237f81a3678fd2ab0e0e1425a2d8bfdaebd0f2314ab95bd43b6ddcde4c74105d,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,87.64,87.6,exact,canonical-source,profile_37e371ca4768fbe805410768,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,f2c19d3ea0ff7daa8ad29ee653f57c22e09d9890f18b0a1c80cda076120cf28b,f255c7983232aa9084d4cf19aad58ccf305681730d842489ef3e8eebaec704bd,f0fbfc9751a9782b3b33aab1bcab998085606bcf9fb6b9aa78136eeeca10f93f,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,65.152,65.2,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,sre-bench,SRE Bench,terminal-agent,a801403c40b5c7aef1c61d7ed547bdf6aaefe42b0ce253fe31c8b757f54991b3,ecc638fab9cee439f89a134c689cedd2db34ada3dd57497d3830325007a70043,4a55027c3948aa511f98b5c335d5a91f898f964d2b20298753f01ca9e7840698,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,33.588,33.6,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-22,fresh,40,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,apex-agents,APEX-Agents,tool-use,2a1c6f5cf8a3f4dadf61fd1058e56f45361c16e1449613c791722864a49f89c8,b7d307a29e8002468892c13578ac62c115860c8cdc45181c767e4d45ef4acb8f,23f10d9bdf1a4d6dc532a2040fad5fc575bf339ecd97d6a824a4fbb34fb2f893,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,73.5,73.5,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,arena-agent,LMArena Agent Arena,multi-turn-agent,f1adeffc9125246a21115cdbb6485535717f91ed02e9821e032a1d6bbaec7ae9,bf2d9775f56feb46bde78f3af9a5bfd65011c105d7734a3496e218e01c236e64,3891df5a797d999b8fe8794b8938e0f59fdef21b32ab320fd7ebd0f9cdcab60e,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.20939685760341706,0.2094,exact,source-specific,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,vending-bench-2,Vending-Bench 2,multi-turn-agent,29ffdce71eadff6f6151fd743dd72a9a42350b584d0e22d79ea6509f370370fc,fa6f4efa1932d9dea8fd06d8d6792cdcdb1f0f54a03203805cd26d499e4f81a6,b9e308c17db60b50ad2b30391faed42194797f8c3a82dca4f1f5b099cdb75f72,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,9235.248333333331,9235.25,exact,canonical-source,profile_3a3f12ef819b20354dad02a8,reported,unknown,,unknown,,false,2026-10-05T10:25:48Z,fresh,26,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,5ca408164414cbf9e4088426eb2f6a44a89af43590326427f18ea98b56085637,7c00f6d125828882478afc60881d59f7fbde5979559c6db45432d9e7f4b2155e,241513c18e44140122a191607f17fc9dd2c39d9f45af88ce27c36ce48b2a37a3,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1866.27,1866,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-08T22:23:24Z,fresh,23,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,f57c79054ff4ab03f5deed77a317fe490b8fa5a39040d40fb49a24ea79d66257,a0593655cd6b6f707e41587893477e13e8aa39d33fae81c7ed9aabc5a89a9ad4,2d6d4e673a66f9deaf161c441bf0eeeab1ab980cb5b19761491ecc49549b85f3,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,47.143,47.1,exact,canonical-source,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
anthropic/claude-opus-5-5,Claude Opus 5.5,Anthropic,closed,2,automationbench,AutomationBench,tool-use,7c495b38a5c9ad7bfd4753fe8bdcdcafff109b2243f6e0cc4dd832221667c419,8de057bd410d5d9ffdc4d3476a35f657850923e7d181ce26098feb11f4fd882a,bf38f00a53ed41f7fe4d189f8e24016158b36abd18578c7814c55d6b17f3cc8d,release_0e98d905ec617087da8bc4a5,AutomationBench · official leaderboard private set,official-leaderboard-private-set,success_rate,percent,true,40.0,40.0,other-supplemental,source-specific,profile_644045ef3ccf710aa2a97573,max,max,true,fixed_effort,,false,2026-09-22,fresh,40,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,anthropic-opus-5-5-card,Anthropic · Claude Opus 5.5 System Card,https://www-cdn.anthropic.com/fc1b44717c85dc068bc6ba5024219938094694bd/Claude%20Opus%205.5%20System%20Card.pdf
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,epoch-eci,Epoch Capabilities Index,capability-index,4b8a3ae1583aee996daaaf62332f651522719e06aa5072657e42062b0687bebd,aa34960b64066775ad8056b867c2e1906c64770ae4fef1f949e42f5bbed4bb3f,9d37bcc57f33aaa11b5ce79844c8a3be804c34656246ff039d46c4b9376672b9,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,168.11,168.11,exact,canonical-source,profile_fad73ddca55d063974c32c5e,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,e2035d37eecf12bb79b03d144ef9b0646ab10a40cd2427a4a9502ccf9df88959,1e254691d4341d3d5524408a40b477bae62c016e6a109b4f5e92b60f6b684287,e2cd58476c62d709590df128f7e6f309788120d3a857d204e134ff7926fc6e7c,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,50.976519,51.0,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,b360e218054c8eeccf0a39b0d88426a3fa1ecb931391b76acbef5bdebce3049d,0cac11b4c96f9dcf1f45592709433961c7967adff7ede6f5e3b5d629312cf1b1,78d83fbc32e7b318f91156ed47963899689d5387245b7b400d58ffcc26f1b9d5,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,46.246525,46.2,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,simpleqa-verified,SimpleQA Verified,factuality,98707626e65cfe1e239493ebc56900a13614bec6a9e27d6430ca8418ad5c0426,f8f4497949fdd184b8c6760e812f6dccb2c1b695f71b89e8c4ff15ec2d0ebb68,ef6c4f75d3d21789ddb6c88f010f9bb1d237f50fd153ff31736b885e1189bf67,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,57.3,57.3,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-26T20:00:37.000Z,fresh,5,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,gpqa-diamond,GPQA Diamond,science,42ac047d1577f4c173cd5f3ce295514e8adac126aa2fb639a0b1676953910fd8,f5dd0bf1f989e8357b1a1522eed176feb5109393daad95d452afd9b468a8fca1,ae5a92fcd50d67c04e43cc9dfcdc064be19627ff985b9e4fb01f50e9edd1e4d9,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,94.31818181818183,94.3,other-supplemental,supplemental-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-26T20:00:37.000Z,fresh,5,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,fc8b5f1c591d90ad7cf10828d4c74dbc2cc9b0b6ccb3159a3a04e8824dd47273,3c08b444337e4391b44c6c6f56e4cd6422a604266e7f6e16e41713f09e795c0f,a3db0f25c15dd80f8f1df610686cf08d4d7f681733644a3036d4c846bd1bb05b,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,100.0,100.0,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-26T20:00:37.000Z,fresh,5,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,frontiermath,FrontierMath T1–3,mathematics,01025299746a32bccae6ca2c95026ba3164866885364f316e43511d6f22b51d3,b63903e3688dcd7419a45ed333ee26c800d2aceed4aafac7879ecddb186475de,f5cab8d251dff3632f3f2c87610226684f9bfa0528ddb04c22cddec3c4a0ae89,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,90.87719298245615,90.9,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-26T20:00:37.000Z,fresh,5,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,mmmu-pro,MMMU-Pro,multimodal,3073e2d26a38c6a5f34d5b38e9d7949cb754421a9dc55068db9960cd59ff9324,2ed2d0b4daedaa368cecac9170c2e6ee24440fc463eac34dffa3d8984f67af48,e91ead25b62406b52b2af4bf8db48d0a75455b005217ebd701e5078a505c1fca,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,84.682081,84.7,other-supplemental,supplemental-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,critpt,CritPt,science,7fd7ac99a188957c65b18d9d7dd36acaff5937d98a5111f704b2e83416d967ba,f712039d59462b86bd57dfafcab59ec9a0196105ead33fcccf5544c2e79c6bae,78f9e5d2623fa0c0df7ebb66e9d9fa411b1e744757f72a9066cfee6e42522057,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,28.0,28.0,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,scicode,SciCode,coding,629a67343581fa25ff20776bd47e10625d338eed4f067c86ff8221b127daa859,e82f6853f4b542eea146da946741a04d117924a87c89d29c36afaa9584474ce6,eecea57ed4fdc58974fe4230b5b34221a300959c5e5c6a80150344fab26542ed,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,55.0925925925926,55.1,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-28T21:25:02Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,ale-bench,ALE-Bench,coding,bd7ce2f252445107005a7008f30b51a6594513a1091a2c57977842c3506d2d5c,b566cac034efe943b6fd21fcbd31e44f9a762886474cbe360cc361d56e466f3b,3f965b4a622570f1f9a2075437a5db4a514450cbdac0122c9910bb000af2db2f,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,2608.82,2608.82,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-28T21:25:04Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,vibe-code-bench,Vibe Code Bench,coding-agent,f4d6c1febb85adc142a91f2cd46db6716b055c23e2e675349948585d2462cf7e,42d42bc3d064a7e97922db8972a24497c58f63ce801c068b3378dab3edde8abd,41e79e2964f6ab021ac4242fa0cb152af108e6a58e724666dae2ef612673913c,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,83.997,84.0,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-27T03:18:56Z,fresh,5,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,webdev-arena,Text Arena · Coding,coding,35f541080c298bd325b7e31a8d583190bb52860bba503287a8fb190661671c0d,a9a2db9175449572100cf67cb852f4db64bcd2bce347c006f0b41c4de891f7b8,f7dc56d40d4260e5791c0016449a4942d23fd3ae92607868c2f304f3114e811a,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1672.6547058209349,1673,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,weirdml-v3,WeirdML v3,coding-agent,5e02557bb2443d43895993337c9186bcf7630d0baa8ceff5277ffa501b23673e,48bcdb7c34051737694f4995e3b87913de5d382533bd0b03d961d3b6f4bbf111,ac684f33d15bf8e92dc72115f8694925aec2eef88d6f0f9917e40d25269b7ff5,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,28.086310681818183,28.1,exact,canonical-source,profile_27776d0f653863ca06f226d9,xhigh,xhigh,true,fixed_effort,,false,2026-10-29T17:25:05Z,fresh,2,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,b505166eb530691643f7b9d782f4784150ec37d38511f0a0a2bc16d3ba81f03b,511b882edee979bd4f47171c205574a0793a289f21aa6295cd006670bb396f9b,cb3e5222b38738b00835f4a4fd4da5e4bee7fa3736fce859b040c4f3c15a998d,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,49.495,49.5,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-27T03:18:55Z,fresh,5,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,apex-agents,APEX-Agents,tool-use,c8785258edef6a28e5c1e3c36f5ebf92454bb2cfd78e5f90a3e19b8e94f67b55,4b7f5fa1dbc173f882b8a9a390097567a144ed5be59c05c2623c10d1591f2575,1df53f9ac83d8582dfe10982c03cc721354a33671800571a398380f102d14684,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,68.3,68.3,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-28T21:25:02Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,arena-agent,LMArena Agent Arena,multi-turn-agent,6be5afdd047ff07c538263aeb91400a37a8bf49f2c8b13fed7f43369e1a79c76,51c8f3fbe4e642e4223a5a3a891b4b402a611283285b57d11068bf39220991ea,01f5243b415dc445a947bfff08782f7990e45798b26ccd0cb03bee3c2281ad68,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.10574985212598499,0.1057,exact,source-specific,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,vending-bench-2,Vending-Bench 2,multi-turn-agent,fc8242e60b385466c5589af9c89e15b77c9305c418c3e7e702fb6dbe834f6d40,133c331a282d0dcf12cc0e2ac4cee13f1ca70236865c16862956240d20226b7a,b239c1b99ae50cb636faa6d17469451a96b4bc9dc248ba558a9dcdfc2672f696,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,9107.458333333334,9107.46,exact,canonical-source,profile_c68c5fd57a2d216f86c6b300,reported,unknown,,unknown,,false,2026-10-29T17:25:04Z,fresh,2,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
openai/gpt-6-1-terra,GPT 6.1 Terra,OpenAI,unknown,3,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,48cf421319097e7b8b0604e431cb62c982f59cc9988f99244fe7222c847a6305,50fa39e17f118f3ee85a2caf2bd447b3dce5f8ea3378fe0149d839ee90cd266b,42b1936483c5fc488f45eb2b6d7373b969098a05bacf7c0e9b60c70d3453c03f,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1475.63,1476,exact,canonical-source,profile_88989cec8b812f0eb187ba3a,max,max,true,fixed_effort,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,epoch-eci,Epoch Capabilities Index,capability-index,ddd0b8257778602f1e00f6d51e0e7b4c94bbcae06e7c7ce0bfaa60344985d0e9,67341918c065b1bcb37fbe61eaa121d61a37b47b5e3306144dc4299966f27429,58c3ce1c018c38dd41fd0511a3491eb91908b10724bc3992a5b438b4628094be,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,167.82,167.82,exact,canonical-source,profile_f344b195b71bb391164d5d96,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,61202f3a1118b82d71e592bcb2f622cd0f1b8fe1614f0d358d8ee944aec66ce0,b8a13d852cb5f0d8c924abb074aba94f685e426d96e598cad1e2909abfab9249,b5ac69a22b19ecaf3c8f59e2bfd80fbc3ab11425ff38427a62926045af452cc4,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,54.508734,54.5,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,arena-text,Arena Text,preference,2221aa322dfd3c62556f67843cb98a432a3006c468b6ff56c292a9cdaa301322,f4bb869031c0f32744611b52574eb66455574f988007629b0c5ec209e00b5695,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1481.3279556386954,1481,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,livebench,LiveBench,general,0d296e2ec3c6a8cccfdce9282771e83865c594c47a3edcf8b5be2ce9908af84c,540319f9cd2912df1ebc6d1937ab8770c99306c932446b17d1d31bb059694d6b,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,81.61566666666667,81.62,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-09-30T18:53:19Z,fresh,31,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,a81cfc7ab685b5e67fc1aa528c1df8a92f309613a5064722f22b0481604d746d,46440da5ae85421e5a475347aeaed5fe756970a6cccd6cd8d85f84b782eda98f,485a315864e86fddb023767c6d4016884ea0cf92ac39c877e9189aa0f652b41d,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,52.91937,52.9,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,simpleqa-verified,SimpleQA Verified,factuality,f2dd5977f5b73df643d315db609c420643bca36ce0975e713bacd9b3afe8150f,322d270eb347c7d03a48f80eaec213674356187082574815c4a44f658c9790d7,f6b3a1b039acb21492a8c84d82bca5974ab8d86acef1fa62417b47cd67aeb27a,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,73.9,73.9,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-09-29T22:00:34.000Z,fresh,32,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,gpqa-diamond,GPQA Diamond,science,02fef9fce5b543000edbd124fe48d58a480bbfbe75196dc083b11693a3d900c4,363713a747ee237e4774b09ae00247b2257c4c959944d72159c9b093c6c0cf36,5bc1057a9127eb8745ece81bd45ce8e12363f10a506b09c4a0e9d6621ed4742b,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,95.39141414141415,95.4,other-supplemental,supplemental-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-09-29T22:00:34.000Z,fresh,32,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,781b25548b23965319b54f0dc5d2e1036aa647f4de199b52cea0c4148fa10072,1833ccdc922c751e1070e5103cc85af255410cad0be6334870ad6e758a1233c9,5b9da07aaafa7046f2395a09b4048ca4669bbb659b767c8dcc651f9e26ec7a95,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,100.0,100.0,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-09-29T22:00:34.000Z,fresh,32,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,frontiermath,FrontierMath T1–3,mathematics,30a05ffc0fce382ccfaa265f7f38b5f99b57e1477fcdbd167cc050cfae90b8e6,b4ed9d4e8fab2f53f15ae67596c0cfc5882d1215092f666334ee6ed032ad1152,702c9408f392e2b8a214bd0c1b07cc3f8e14d785c34bb58e89fff092a08d6449,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,93.6842105263158,93.7,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-09-29T22:00:34.000Z,fresh,32,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,arc-agi-2,ARC-AGI-2,abstraction,df5abfa7bdd40a5503867a372e0c8fd07426aab0231f48720f1313e159ce5daa,85c64eabea4dddba36ea180ffc8b75b5ef01bc8098269c30fcaa900e3cccc788,2921570a0438a4d0f49597cd194c8d87d5de98dc53a452863c498b702c27f139,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,94.16666666666667,94,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-01T00:45:01Z,fresh,31,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,mmmu-pro,MMMU-Pro,multimodal,1e3ab91a70e56bddd692fe4155285bfe5ad4cb667b6d5ebb0a4fbb22f60eb313,3caa28662f526b2d8599f1294fc07f8a1ae99d0bac9f781481113fa5085b5b2a,9921d58127eeaee7b45d9b47f13c5c6fa578e8390356c58a0e2e3e2b423d1c42,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,85.953757,86.0,other-supplemental,supplemental-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,critpt,CritPt,science,54b585f00e1319c168c9ad7757d36ba1806c667b29bf72a57ca371e5495b4b3d,6a55635df4c3bdba50646808f2ee0c2a1799da2ac1bce88199b00551c249fa37,3b319bb2660e1142e8aa68dedf83a1e48cd257c8e82fdfaa3b3d28704f2f94f4,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,31.7142857142857,31.7,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,chess-puzzles,Chess Puzzles,games,9c14f265e6a72a7799e2f206071a46d79c017b84cce897321fd2933b209b1df6,6d48f5aa7a939bede8402e6b9dc42017bb6ce5f5716f73615b437e7086b7dd4c,d29693ae5589cba535a0d3ac93ef0e493c88c462f05f6523e35f7724a5112601,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,68.0,68.0,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-13T12:40:12.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,arc-agi-1,ARC-AGI-1,abstraction,3aa17d51a5d744f6cdd6c9101ee771d01dfb0493fae5b19d6d08663207e0c1b7,a80f9f67dd9d4eaf9be0015eef7c72e0d531eab7aee83139778a3222da001985,5d64108a49bf54a9fc6a3b3849dbe4d14948c397253a818f5ec9b237ebd6a6d3,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,96.5,96.5,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-05T10:25:47Z,fresh,26,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,a43fb7a9d0c719099314e494a12d3d0e5672d1ee6c04cc4d42b3ad1183d3be36,29bd243e5870a3dd6b33af5bd1e599dd623578e67ed39d03e404ec46746c184b,93332fa6cbb00a0372a3c755eda4e5f67f5ade9f21e1321a156d492f77fb5453,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,96.11,96.1,exact,canonical-source,profile_459bcec00071cd59cf8454da,xhigh,xhigh,true,fixed_effort,,false,2026-10-07T01:44:58Z,fresh,25,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,frontiercode,FrontierCode 1.1 · Main,coding-agent,0d8191da8224992f5f8cb4d65bb6d2f27ec4f4af5e641ace838db64d9c095055,38569a3c91d174208e19a8061c505abd58b3946f364a068f87c017f3bd9ff3db,a162760f44520a96a355010193409303db46962a0d7c9a9d64a0e8c05c69da1e,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,50.23,50.2,exact,canonical-source,profile_64133a3ee4324a3d4e8eabda,medium,medium,true,fixed_effort,,false,2026-10-01T00:44:56Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,ioi,IOI,coding,ac1ad386b9334cccbfa659158bc6ef6e6fd111200b2bfe2441318d073f7296dd,b7b220359511c0fe8292ce5b4ac9ea2bfa23b0364a149e7184f5c05595a07d42,bd1e1840c14c84d69971323ad13972c08dd9719d4b25701f7af54ad9f2c5e5e2,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,96.889,96.9,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-09-30T18:53:22Z,fresh,31,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,scicode,SciCode,coding,974e0dec0661904b54a5e9897bf739176161a8541e84c555417b5dd29ad916da,c3273d1ac5304520aa6aaf7ad13c9a84ee06ce979b6444dca34277bdb70d4a38,0b83a8eb7770cd577823d10bd155718fbdcaa8a113699953e15607a69cb8adc3,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,54.1666666666667,54.2,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,ale-bench,ALE-Bench,coding,0deccede355734801a5179354ec936e3b6168a32a7888b8b4f34494fe384b4c7,b0062de33b311a624d070229cd30cbb861650ac79b3c990ea054876d35198806,a6ecf64ee9aeda07ecff54b9c2e22e3a2e70e22b41b4786096462fe7bc5980b0,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,2743.75,2743.75,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-14T07:48:15Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,vibe-code-bench,Vibe Code Bench,coding-agent,538d59f6221a23388ff97077a4577a9dec797f329f3eed4bf3528d8cfdaaf56f,39d2965f3c737993ecf8fe9c96769dae1d5f437b0004bdb29796b9c949b316ac,4973493a3ee4b091413e398b03d915d192a4ab752497d5231d7ee119e00d73d1,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,88.931,88.9,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-09-30T18:53:22Z,fresh,31,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,webdev-arena,Text Arena · Coding,coding,b39bf680dc05ec48b4807d695fa847874ae1f7b6c48788ff69fed124f29d4e76,0ff905587ff975b45288ad36ec5d3984c60d61f9175b3a30ecc4795e9b586d5d,a0e12b15d7f0ba66d24db72fe1481e4959c68e3ffbc550105c7d4f27eee1a02f,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1750.568215754016,1751,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,mirrorcode,MirrorCode,coding-agent,c0440234f551ecf8be7bfb951f07cbff7d1a8af1ae5f4751ab05304324d93e32,bee2ce578d192cdeaf841cc552636205b56f0720b8b7192812fb6b4bc462a464,5d881538bdfbf8798e04801451260b635fbe881029d76cf3e232f00b690eb88f,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,61.111111111111114,61.1,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-05T18:36:30.000Z,fresh,26,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,programbench,ProgramBench,coding-agent,5037646c2374de70feeb79406e431ae3ab89a019d70fa375f9ba94cc5c113ae8,1c8af16e44bbb69bdfb45d75b91661c82514ccf914530bf6ec8c63e0d36474a9,9d064e8c7fb4d86c8c99771f88cc81639f127d01ebce2af8f1ac720e8b5d3c28,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,3.0,3.0,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,weirdml-v3,WeirdML v3,coding-agent,c8a5a2bb1fd19e7239e259bcacf5768cc990e4d186c50435e7ffdfbe6337782a,f7dc18ae893d3bfeb2b97eb3e08517ca78628ebba89ae8b88daf248fd6e085b2,686c8f8848fdd05629f4c1d78a419a7e068546e2456745fb63fb878603cb7f29,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,38.25364136363636,38.3,exact,canonical-source,profile_459bcec00071cd59cf8454da,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,2369a44e333a694b1077f238307bcf8c50fe5070a83e16703119249624e51a9a,36d63a23f595004e87141d8db9d1cb74f8d46c4e8cf73e7a6827b7c177432c1e,6afe21581a06513b462b8119e0c1b4468ede7836d53b0e04f2144bdcf701b0d3,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,55.051,55.1,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,sre-bench,SRE Bench,terminal-agent,7b0ec6febd86a525d5fbf9c0e2fd0967498e545a59b6ac69904c31e6b2fccfc4,fd7fa5253eb6ac797829361f671b9701bf8783b4863785c6c08663baea3ee5e0,7d2a28ed00d6ee5b747a79d0948a58930c2238d3e0c93f1a4b49e4ff84353eef,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,50.763,50.8,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,apex-agents,APEX-Agents,tool-use,b01b33724562c0286fa5472634fddca4cd4bbc683d29d379eb75a7fa3855d4f9,ce1318e9329a8b958f706b9305f8f005548c950e7fceaae2bf4040ba1344e9ef,689ad11ccd00a883405b1a07ff7fc60580154bd6734b5797405b32320d760c7e,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,60.0,60.0,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-01T00:44:56Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,arena-agent,LMArena Agent Arena,multi-turn-agent,c9bf739492d63b0b37385677093c2ab362f2e0e11ba580d7a833b3bcf0c0c414,89f47e8a7a683995589c62fef33ab70cdc3d2948f9a3afd27308530a1e9ca1ef,2a681a457c208dcd574e1bb7305ab92bb199f91d7a120c9c5c0e9000ed808849,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.12466693893518165,0.1247,exact,source-specific,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,eb0491926d90b0813a9ec5fffd6cf863c893eac16f43dcd4358ab567223e5070,5f96da894d2ed8ab3a66a1eaa85b96f885f6dd1ebff40ac96f6a4a8d3989dcca,d757e6a98dac42027709d8e0a6ac15ce67c2dee656acd2b19cb0474732f791cc,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1608.89,1609,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-6-1-sol,GPT 6.1 Sol,OpenAI,closed,4,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,225aa22400e05396bcb6c788fe4c2789f11746a42b92e1f11254c9694b45d8c1,d872d3a00cdd8eca8a0b47d44e642bac26ee95e55a992ba9dd4f54eefdf97aec,0f8ce86fa909356964117c1295b0616b2f5856795f89c8e261aaa8a91095728e,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,52.857,52.9,exact,canonical-source,profile_4104ef8cd84a9215477511fe,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,epoch-eci,Epoch Capabilities Index,capability-index,558fa012631345be5d353a741f3f449538c1336af7dddb421cd99420adba25a6,9b0a7d78d3e795253799a88d1dc41b347f562b84824dc680c9ca58cf33afdedb,143062028388ab8f96c0a8e313d95fd93babcb5408afca71858f88353a6be75e,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,166.74,166.74,exact,canonical-source,profile_79445b70144959c79964f01f,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,6cb5605778e50c78e1be5f11ad04f85cd35bb4b76ab59778b9417e434d278500,f0871bb220ab40f9ad0e37064196584be97a257cb0c272525e42f88d4cf42175,1c775fc17e5b9f8df64968c1f24fe97a90028397c48b50a2dd3880e26b87eae1,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,54.241334,54.2,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,arena-text,Arena Text,preference,70a60c175a36a813387d408f0e65a87078889cdac6dca731382a06ba99090d5d,ddd252fdbd9fd816ab8185813c992a3f0bf6d30fa8a86b9abdef8042967a0394,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1477.1495460775536,1477,exact,canonical-source,profile_c047341bcdf46221302633c9,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,livebench,LiveBench,general,561de4fc8e94bacac3cd657eda4b8f1aa7db0247922ea3bcb7b246ef0b236cbd,a7b05f561a989f788466b3093a8ee4c8f5051fd43244727adb6de5d45f159841,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,75.67214285714284,75.67,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-09-30T18:53:19Z,fresh,31,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,c1b012ed99a9868ef18c4775ced59f99d6486dca5bc9a694316128ee66c4ca55,fc31a5beef0a3a5cdb1973be4ada74d475c2e46dee7d2006c573ed4f2cc7db37,58c4cc30c816125af54eae4d3abd0a331486129f2fbc5ff3d15a80fd0047d3d6,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,46.12,46.12,exact,canonical-source,profile_c047341bcdf46221302633c9,xhigh,xhigh,true,fixed_effort,,false,2026-10-12T23:23:20Z,fresh,19,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,e1c6716641f0656ca993e8f0bb91c26bfc6fca6bf29b9f2bd3ccf9c6823eb56f,83e9b61598260144d1969fa1b07d3ffa313593f7c701d582579f1b942361b6d1,340e2cd39df0fccda00b3edaa8ad0170939d4b185bc6d2fb1a80d184da0eee7d,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,54.958295,55.0,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-08T22:23:24Z,fresh,23,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,simpleqa-verified,SimpleQA Verified,factuality,aa5fdf947fd38c12937cf127c39e47dbf772c7b3154ed51f66caf80c99bf1e21,a86e3ee068926c16c8bb2c3f8f345d434da29eb250cc8b33d9c7d162ff7b1b42,24eec68deaf023f9661f8f974a8fe5deb3efb84f49f19ed44801de89e1b56910,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,46.5,46.5,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-09-29T15:15:14.000Z,fresh,32,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,gpqa-diamond,GPQA Diamond,science,b6b45dfa195b66274c167f7fbbc2575d0dbc24ef150d24e43dfb0ceca6c28ea9,d7098d469e5249bd3e9b6fe35ed6e9d4d9d5b716382bab4caa0e6ffb52ad178b,4a1225d698508261ad8c8935af5017f4596cb88e40c0d3222b44ae5f62fe27ed,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,95.58080808080808,95.6,other-supplemental,supplemental-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-09-29T15:15:14.000Z,fresh,32,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,38eeca962257c1e216b4313cc6e4ea0b29a15718f100af5b90ca68272f842515,7772d7d04af6f74decee8cddb88de41ec4ea7b2708d7ce8bd65954e578bb10a7,edce3ee4cd626f35a9389beb93d9db7f14028e8cd382a514775b7b542e2944e5,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,100.0,100.0,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-09-29T15:15:14.000Z,fresh,32,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,frontiermath,FrontierMath T1–3,mathematics,f888cf48c322de468cda21e087877d752687923bc04f494f5b6c550f794252c8,c3e84c26146933e092886d23478ddc096e9a38f565a21ca7d5689dbc6d55c530,3be3e4d5ae97d1beafca7a96a0b7ad657325b3715a1928728c8c592421c78f2a,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,88.7719298245614,88.8,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-09-29T15:15:14.000Z,fresh,32,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,arc-agi-2,ARC-AGI-2,abstraction,a07f6b0dc7f91f5bd0f3985f623a56fe2a84a80d4bcb1a18ef73ab1e7931b587,b72d6a3c7f0a1df76de7dcf19ce05730291c9a0dfcb898797020f8404352e36a,b107ba3fb62ce257bd87b3ddea05d727353371665b5fb97a169730e8c8bb8eb7,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,83.88888888888889,84,exact,canonical-source,profile_c047341bcdf46221302633c9,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T22:23:16Z,fresh,10,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,critpt,CritPt,science,26401a2d36607480fae57e101c10a998fb727967046f3646740be75265c8efbc,63868f49baf03d3698ac2ada220b32c103d6b1e1b9a339815e5e574cffad7ce2,9a8e6be9552e6c546fc2d64bf9c242d863e2ef802325e92e090bbadf9ea86d55,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,31.4285714285714,31.4,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-28T21:25:03Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,chess-puzzles,Chess Puzzles,games,77a3e6db92c2bcd1c69ebe8d0149a16fb4ced19a7cf6bfdd09c4b9626bed3f2b,9131b8c10c0a2af295b692eaac110d0e654c804111f7c7672a209262d2aea778,561b8c04383c8f9c20dda019b47f055033234ca9a8af2a2d2b0ff1c1c79fd3b4,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,59.0,59.0,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-13T13:00:38.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,arc-agi-1,ARC-AGI-1,abstraction,5e5b73084b7d6dbb8356eddac1a6047454a90c35e9e6e7316cf26e376fc6c6eb,3cb99f97ced83182fdacd242bc7e01b4c6b9100852d6c67d54b03ccccc7705da,11aeb5821ef99877ee2e091a89e9596251f22782d05b6fcb012811273f71bcd1,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,96.5,96.5,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-28T21:25:03Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,f949c7ee783bbb0bbdb8b127ffe147a9b79674685d0fdc0245b391a077a5900e,ecd8f8c7ea9daf3e7d8fd0e384f00630aacd7bf3d53a1709fa5797f089c75a0a,3f26ad54b9ef5b07a2431a1788398ded49d2173bf40feceb00296d2147548523,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,95.79,95.8,exact,canonical-source,profile_c047341bcdf46221302633c9,xhigh,xhigh,true,fixed_effort,,false,2026-10-12T23:21:32Z,fresh,19,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,cursorbench,CursorBench,coding-agent,76cf8b98dccbee48a37b96687e344ca8a992af11824feccc29f8078c6e855be7,4da71e44e75fe263a8f77bfb48641d0489b2a22e197a013ed0505c2c529e5768,425ef424c3cd3ddba464844fdcca8a9ab839a0a814d8fbeeb9735c6d71015b03,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,55.50000000000001,55.5,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,frontiercode,FrontierCode 1.1 · Main,coding-agent,4177309dcf39ddad3ceb77cdd62ba57b6095de3ea77ba1ccc3e1fff9e7abd040,c3974abb3f8a6e20e6f4bbc27c687d15c386d26979ec0bfadba794d97a4c176b,3d7c5cfb1c3a83f418c179444db1a506184963910971b39d21696da7073e9e94,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,52.09,52.1,exact,canonical-source,profile_c047341bcdf46221302633c9,xhigh,xhigh,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,ioi,IOI,coding,6255606218746ac71533938cc5801309585d9be266c34d698d000324bc405cc9,4e3abb35897aa4d3f0df79e07d9ae9150ac4c66b3593c89a79f2300cdc2643bc,1b78bf03f5ba3ed0c2a91a378163858cac3be88338879397dd5b6d36ca883bc8,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,83.056,83.1,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,scicode,SciCode,coding,e0d46942e3465858a3887610bc835c44f2326cdd169f5f1371b89eebedab4231,289fce8c377deb8cbac33ee93f17df7bde743e18a403aba50edfcf5d3d8918f8,3077b19518f6047f5119eebe79fb42140c7e78c4b0406c99a125fd532354b80b,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,60.9953703703704,61.0,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-28T21:25:02Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,ale-bench,ALE-Bench,coding,5d3459670495c89ecba8015664a8ed56a3e71d5e22604405452ee1e263580cd1,88cfcac96c3ec489d1e8fc9b5a87a64fbb875544c0190f525922b8e72afa1698,720c31a81a42023e24e49c35e2b042b618ac030f394a64312e7e7e475578567d,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1819.12,1819.12,exact,canonical-source,profile_58c2b77aab1a8f5d917ece79,high,high,true,fixed_effort,,false,2026-09-30T18:53:16Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,vibe-code-bench,Vibe Code Bench,coding-agent,463ac7607438a0a11499b28ecce7601030614704ca028c7eeb2cf733a83b81b3,644ceb16faf9e402ae20ce04b4cebc2820a6248a0ff6c735e8f7926288ef347e,df91ce5a6e052678fa8c25a5e100bfc2d5500eb86f692eaaf06bcac40f00d6be,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,92.392,92.4,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,webdev-arena,Text Arena · Coding,coding,541c77f238eb13105ededdc974105de5eca21014c536402eed2cd68ad82e0336,46fd331b4a8407ceb52fffedff203f621a0e3c674f45b77e427f0ef78bbd6673,7a5ee389c0130c4132b44ac64700d06daa93f4390a593b23574f490f39b02d85,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1769.7283971913196,1770,exact,canonical-source,profile_c047341bcdf46221302633c9,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,mirrorcode,MirrorCode,coding-agent,487c6e4a443aa00b259804e2baa9a677779803a42a0b796412808e616fe5c54c,db14c60fdf03a970538c98a28839c6b261212558a2a37c62799564cff900ab91,4d416c3d3eb64050c2c82e2c58eb95ebd78f89712a32e76599eb45e81341dff1,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,97.77777777777777,97.8,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-05T09:05:15.000Z,fresh,26,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,programbench,ProgramBench,coding-agent,20e362401668ec7b12360d4627efdcda91a436e05d2284395287aa27bdad5a5b,777fef7ceb238f2cd4026cfd8bb67f62ad7f45ebb6c60eca1dc309542905bce4,86e7e097653d8a6be93c6bbdf319e275034fdeffda75980c374006c9d747dba5,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,6.5,6.5,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,weirdml-v3,WeirdML v3,coding-agent,c150d5d6b5b1ad68e60c9a86d0838020d21748db3b11ed95b9b37ceffb7d2c3f,159c1e286f6e11fb70bf90a1824300d9481434058ec99238a55d19a725bd04b1,c5065a9004b8597251058508bf51d33f3ca9640f5b95529999482ae6f8c9a673,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,19.940320346320345,19.9,exact,canonical-source,profile_c047341bcdf46221302633c9,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,f59849db2ae3ae18bb354f9d333fcc9f88bf5254a085180051bc3303d56069f8,3040c62d27cbaa1dab54e8a36b84352afbebad674b89ee6f1ca14757394d0998,c0e27a13632b8492ee11b340f2dc6f72361f6528d66020a33218c0534e02175e,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,83.146,83.1,exact,canonical-source,profile_58c2b77aab1a8f5d917ece79,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,7e9d00a7bfbfc9d494a50976b2d7aeced73a2093132b95b6528585dc06a3d45b,ba8b517b5ab1e26a293c82db547775861c6df784441432262c6745c024e72dd0,1e968cdb859adb68b42754ff99c3d73a1263ddc8eb1751f5d44da031b34a8c68,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,64.141,64.1,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,sre-bench,SRE Bench,terminal-agent,3b86602123b277b25f40dc8aa5aee4c39febb3fba6c5d75a03a93662e059dc49,fa5b138af7eaa464c0bd8870b2761f6fea490384757208a00529a3a8cd2bc1d0,ed74b9ef8fb58928ee988e2ed43ea06d08fce4c23a6f9879dca908dcbf27916a,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,30.153,30.2,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-09-28,fresh,34,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,apex-agents,APEX-Agents,tool-use,77e129d73ab12f49050fa06cad811a14fc8cd5263882966a6094965eb35129a8,169edbef85191bc3aaa57bb82084feb3bb2bc7d85174f19327038013959e6498,ab8e1c920a10c544e177e19708ef963e04ecb54fb4b6b66f7e682ada59371f5e,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,75.5,75.5,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,arena-agent,LMArena Agent Arena,multi-turn-agent,802b776037082ce7e32a945717ff2d556acfe555540f56597348e3daa669743c,5c1955a8f0caf4f0312fd184bfb31efe21a249b43f32f8c33ee0a2ade87afc6d,5c15887b2bb803c0820a3cdad93c513eb24924b07f6b665cb2ba3a916d44e298,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.1383722472009646,0.1384,exact,source-specific,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,vending-bench-2,Vending-Bench 2,multi-turn-agent,70e276184e1e0e0cd03fb9727151c4245cf815718c4b972736ad36bea62a76d4,f1b677217f39d2cf03b970674a22655c22b3ac9212b5fbf0c66d825b20aa4301,e98054f2aa4577339574ab6d680a555bd1279f1ea4f32a37690727aa0ba9c9d7,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,11165.27833333333,11165.28,exact,canonical-source,profile_9ed9bca0cc3f77c04837bf35,reported,unknown,,unknown,,false,2026-10-14T07:48:15Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,21779bbf3799a59ea33113b39c5601977fbf848e712fef82cbc98a19f0a3fb2e,422c113f97521fb3642a9bb82d534fbaa0a2516886742e74cf5a92c44c7a0a0a,daf855b28ed3d26531ed17deebddeebea38317b0c8247cb9f3023251a6b44a4b,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1837.82,1838,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-09T20:33:50Z,fresh,22,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
anthropic/claude-sonnet-5-5,Claude Sonnet 5.5,Anthropic,closed,5,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,9222cd585cde507f32997000379dc9f217bb189541efb6e54f9edd42ffd4d91c,7e6fdb87350f6587013574c6d01359f6c24432dada5cf13f64971643d8dc9105,add3474fde34ea3212b531a0369c0b154c3e8c1e46dfe972c36d3d0651fe9bb0,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,47.143,47.1,exact,canonical-source,profile_fa49f2bcbf15eab57b062ee7,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,epoch-eci,Epoch Capabilities Index,capability-index,3548e9f928d8eb7e262ca19ebe735690d108ebca6234f7367dc89b5079217282,f4d25126f0516d28dfb6d06bb40f1e688b635acde7d12b78ac8fde018b5288a9,4e9a33cd9422610f148c4d65f3da9d59dcf1ff9837ec4e4559aec3564a6e2a66,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,166.36,166.36,exact,canonical-source,profile_2d5b4becde2ddac8a06f851c,reported,unknown,true,adaptive,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,577c81c91c1e32be33956e1d72c89ca7f29ed105cab5f8910257a77ab2674644,986f273b28652fb06f4eacfce1bbaf7298e0e089deff602d2001cbd172fa5fa2,87fd8b1d9d72a3f67a7d4be4fa0b5c63ea0f6514af7d98e7de1c5cee1db5082f,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,49.010212,49.0,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,arena-text,Arena Text,preference,7c6f0a28f5a90914e505ea3c45c810c6b9b1195e6d3f2061adc87329fc407a44,3a35771755e922634aa961f17d716876582abb88c862d3ac6fa8761e1801273f,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1504.3663236556554,1504,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,livebench,LiveBench,general,e9da22bd58d0345711d0ef85f1a7fd9e047221969180ada297613e2ac8aaf9e8,9946517e5d4d581247fea5f632168da1099e0006ba711bbb632c68d97e634049,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,83.41415476190477,83.41,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-01T18:28:14Z,fresh,60,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,47d11389591c40e634b8ef9c6db794735e6ebbb4eb602d8bc92f89b42aebbc06,6819abf207e2095021e4570fea1b2df09d28fbe7baf023c1b0683caaa6a91273,8be9b8a446f22198220feb716182be194ad5c629dc6d0915009774f26510cbe1,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,46.5,46.50,exact,canonical-source,profile_b672f752d3f34b6d6a8558f1,xhigh,xhigh,true,adaptive,,false,2026-09-03T18:22:21Z,fresh,58,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,3328c8eb28ab27ae020a8033c1f99cea5e65036a91bde15d857ec8d5a859dcc7,1b4b9324ddc872ebe531f791d5f2caeed657c2430c46fdaed33b308d35afec6f,268259924e8b82cb39eb317c915a9f410a0e2a8cd02195f1d809716bf9ecabaa,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,59.128823,59.1,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-08T22:23:24Z,fresh,23,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,simpleqa-verified,SimpleQA Verified,factuality,761aa2b0231e7c595e67dff9650b4de6fd616e1ae4620c094452e790612f98c9,aa1b6186968a2ec8b69cf1561bf0cfb04f2674b48d7057ef2286b006c7f6725c,37bef27d5bdce691247742e00286686d38e2dcdb9b6fb961c64286a4c91aa07f,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,70.8,70.8,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-01T18:31:59.000Z,fresh,60,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,gpqa-diamond,GPQA Diamond,science,4c392182efb325adb9a30008bd1484c3f30dfad2848298e43d04e2ad4442a0fb,e8d047e58f339cd1843d07d1f511a2155deb26bedefc2484f6a5a741010ee169,f755439a1549e2e0786a24466e9d9a05274ed6e379fc8ca962aaa8ec853c2cc6,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,93.434,93.43,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-01T18:28:15Z,fresh,60,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,mmlu-pro,MMLU-Pro,knowledge,d9e6ede5b341d1d59b39b4b90d0b006c79456a2d0c4057108060473940023b60,d4e8ba742d8abb62a9ff2fe70f728683a8768b0e9b22ca1738ca284fdb4b167f,a0e307527a063cc4ec26bc99b0cfa74b7d20cc8e610f8959a86a1e57ba472327,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,92.375,92.4,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-01,fresh,61,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,e79fa1b39d5ad6d5d1f466c137ffbafe318aa1156eaffd2fa6cca04887cf8308,3cb7cb5ac298311170f22ff04d187349f5003bb7561f00266f84b7091561da8e,db287715e5e4ff598049f787f8ce4dd10a0501eded5d1d5d2ab14adecb1d11c7,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,100.0,100.0,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-01T18:32:00.000Z,fresh,60,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,frontiermath,FrontierMath T1–3,mathematics,d90a16dc51340c3dcb813435716ef2051e9aa6c0d56204b231bdb91308ec319d,2ab32cb0f74a50ed7f63b9c026b32093f962de7ee39150b9d8b06bc5bb5fa60b,14bcd4f8da0ac33a2bb0638ecd047edaa58024aab82dbe53bdcdf9bea305f45d,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,90.17543859649123,90.2,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-01T18:38:24.000Z,fresh,60,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,arc-agi-2,ARC-AGI-2,abstraction,66995baf5ed6aeaa4ffcf4a067c0da5b2fae86ee747a915623cfbf2358eea0ea,2031f4add5e65b57bc2c7f1f96df8b1dd12602d9034d008b08bc57d36b6581cd,7fc618538143bfc630a5e98873be0750b7be8cd7799576d0787347fbd031e46a,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,90.0,90,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-02T13:33:02Z,fresh,59,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,frontiermath-tier-4,FrontierMath T4,mathematics,dd35369376d7825353f6e1ddcba4c9a09ce0805dbfc7c0c41a961bb393d07101,fc23b963a297744139396bb4df11bcb3865044c11403ac7218c0d341baa57d4a,fcd0c753c48d67aa1457fb58f84ecd176bf9f909c0168f7ca1082626781e83e9,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,87.8048780487805,87.8,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-01T18:38:24.000Z,fresh,60,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,mmmu-pro,MMMU-Pro,multimodal,e4170000cc9db5d08073196432881709b0799d09d03e7dbb42208ee1e4e4fc27,f8a98f5df3eb70c877f887ea2079bbc0d878b5329028974decbea1a428471e41,c8cea3dbb47b767bdf9177b5440f57e840833e96c49b383ed38c82afc74ea195,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,90.636,90.6,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-01,fresh,61,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,critpt,CritPt,science,d8e1d9763fc874b4df67f1b160c857e9ae386f98b2d85abc028a090a73c8d87f,1b4ab72bf34fc8571a8d3586b386edad8af882c9972ed1206bec1d9d04bbf7dd,cabb65b1023c8688bc23d3eb2feb7764697e7a39921edf05375f613c9c28c7ce,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,29.714285714285698,29.7,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-28T21:25:03Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,chess-puzzles,Chess Puzzles,games,b8c66e946c372f5929c4c683ba96757997328157813af7740ec4f63df5e2f53e,38cd546fe8ff19a5d348067ee9d22c77463c4b202117be1600f8db7f99d7cd58,a4ec588ce8a1d30d31b460ed1fa76cdf9e7512ebfab42302ba3ddbcd32da5497,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,48.0,48.0,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-13T13:13:06.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,arc-agi-1,ARC-AGI-1,abstraction,063c9e6403aae9c56607251e619768bffe62c5d2e219b42817dc81a1d9f09cda,3944401d9acd344152aac7d25680e2dd155486d13d3f4b1a8190567d2d0dff46,698663a4078ad821dc00c6db456b1c588d0dd1db5c52ee8a1b5aeb8dee0a4397,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,97.5,97.5,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-02T21:27:12Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,78cfcf9bda26c31d56b8150d5d634efc58ccdd96c882b1d7b325bcd69370240e,2996be163456988e515d665dde7246edf0809ab5a2785a822d9c280918796840,78a93b3d6982c7b71cbed058bb054a9b31551d9f97d6554eb3b423cfb691ce9c,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,99.1,99.1,exact,canonical-source,profile_0e5a1af41c411d83363a781e,high,high,true,adaptive,,true,2026-10-07T01:44:58Z,fresh,25,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,cursorbench,CursorBench,coding-agent,6b2ed94396fb335fb813f1947c8318b745d89679930a72a014b4c6823445971d,38cec6c9a115454dd03e12ba449779a115cc11b6ede612c827077dd4354faea4,8b82e753b57fa6a00edfa32784004eebb2f66a80416a7353dadae9424bdaa4c9,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,51.800000000000004,51.8,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,gso,GSO,coding-agent,7a2c57c3934fcace58a065231da88061108f198855dba21357ccb8e1d673a118,91151bfe520bd6f00db659d2d1c3a4d34fb827b67de1921159299ebd844e5fe6,3efc26960e1fcf1d87d9edfa0844ceccab8f2b5b2ad359b68443100de02fe22c,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,88.24,88.2,exact,canonical-source,profile_7f837c92d0ca5e97d510d67f,reported,unknown,true,adaptive,,false,2026-09-27,fresh,35,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,frontiercode,FrontierCode 1.1 · Main,coding-agent,28c7e8269b0084a1e83d9406929b323d26707ff16bdac277608a21d7025504c9,7408df9d49628bdec0fa4eb1446998dc32842635ef80fe511f77170a2eecc9ec,e8a4dc91b42d7f893f4a18b6b6fe531a9dfc9da8d9120206ba12d18a3130af04,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,50.91,50.9,exact,canonical-source,profile_74602224514915764af40888,medium,medium,true,adaptive,,false,2026-09-02T21:26:42Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,ioi,IOI,coding,a577001c65c5dfda1807f39a22af204a9078438760537591e524b30b8fbef87f,ec843d0223bc4f2342a9750022ed143c3d7eeca66da07bcb68f66b3a0e8faeef,40886ec6d0fa07f0512adb518ce93f7a3b2bac4d510a7b22739b0952f5b78afb,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,90.778,90.8,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,livecodebench,LiveCodeBench,coding,5d7b51294bf599b918b0135006efc66d053959470e3b36d7e4fd1c6358cffb00,82ea751f1b30c25dd24fe8405e4f8b930866982f8b63cb21a66ec79387ea900f,8a37c1c62b912038268b5e5af08ff7459a8ad551e3f1bb0b1a87256419ff6036,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,90.515,90.5,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-01,fresh,61,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,scicode,SciCode,coding,4a6ee64b57048502a8b7a7fea33c6e30d9abad6118486b0ad6bbd53fd0f0951a,bdb055ffe320a5bb730e82043d5da2138814a17df1ed6b1a6ad1163910de9fbd,20de680c4bb29e88259c541f2c32c497a4ab76de8e9c3539720cdbe55b2d3d2d,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,63.078703703703695,63.1,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-28T21:25:02Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,ale-bench,ALE-Bench,coding,b91224ec986a5e653c62ec511402e02ffad5b8a762a20e4ca0dbf83a3fc45447,8935d305d67eb14ecaef3e0aec77cf89ef203aa54b4051da2f5a4fb96642472e,f88ae9977055b650bebbd66d326e915ad29134999b41ffa6ef9db480a625e519,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,2143.2,2143.20,exact,canonical-source,profile_0e5a1af41c411d83363a781e,high,high,true,adaptive,,true,2026-09-06T19:02:23Z,fresh,55,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,vibe-code-bench,Vibe Code Bench,coding-agent,1bffdf9585757ec98beac2a456fde7454fdd4219c9f51baeed159f518147f7d2,80ee0e6c6fa8f8a3dc773449506fad6f0c65b05ea99d295cd0d4a74cfd4562ce,0a192c199c16bdf84dc145dcfeded1fca7f25ff7545ab24df4daf10450a7c7a2,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,90.263,90.3,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,webdev-arena,Text Arena · Coding,coding,7e719c59b3dd54c0ba98a029fef64310da9a8a20fa797f4a54e3b67d8247a4c6,77a9244958d5d20c0e08c98fe95c9af23a95107ffe1c75bfeafc550d0e489063,a6eff4345d48fc128fab6d5936509f95eb3d137859284fafb04ac71b5aa932bc,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1745.8518267992542,1746,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,mirrorcode,MirrorCode,coding-agent,2fbfbc2ce603c0d9229c876381675c0ddbd32f264d258a1c0b2d06758550e8d9,a1699e79fe31ad5a30f695d7b4bb5af385378b08bc7cb6413e9c01917dc534eb,8fbaf6a169d62a86d13e6fe21608cc208bcc7833433bea2f38afdd0ad3ac1256,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,93.33333333333333,93.3,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-09T19:09:49.777Z,fresh,52,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,programbench,ProgramBench,coding-agent,88a4165cc2e382fba472927459b57a0626f06555aea70e3c9181e6f4eaf7f411,d2825233005db72268aa05dd458eaf18e146e26f3705cf232f0aceca1065bb6a,afcf365e679f6bae5a3db004db3cb7d5bae11d1a9eb0194b03dd93bd1571ed5f,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,7.0,7.0,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,weirdml-v3,WeirdML v3,coding-agent,7460ea7c37a0c6be278946998df9725859810c0c1a53fcc3620402c0161195ce,5921c69dab78065741901cadbf1e7feda41d7722494bc4b444e3cba3ea96e9ec,81824f0520d15d25a22527585671a8b99abc17d9dc2aedd724574663c6a0ffff,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,25.970212727272727,26.0,exact,canonical-source,profile_b672f752d3f34b6d6a8558f1,xhigh,xhigh,true,adaptive,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,f664f74732324d8d6b3b6b24023c427f88bc5971118eafb2ee808fe0d628f641,5f713e0cadc1a23ee21e89e4e9df1013adbf6027fda9ab7e757e640b2c56741f,6929623740bf0e9f6d86bbcc09d072f9dec734e5a138d1d19191baa09fe5ed1c,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,85.019,85.0,exact,canonical-source,profile_0e5a1af41c411d83363a781e,high,high,true,adaptive,,true,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,70ee188a2b5fb5c100485d0ae5d2f25a8ea2461de4de9135f463b1e8519743b3,12281a418756d65494aadf30cb5261867e40d5b4326a4209b7b8ef95e7152eb3,b15767a450642a13bdb290399c737c70043262dac841ba4553929159447b6c7d,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,58.081,58.1,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,sre-bench,SRE Bench,terminal-agent,c9fd5fb69116cc9472a4d297053da2811fb4e994389a7211cc655432ce270ff6,51e9c55c19d0ae46b09958055e9570e17e86f2947fa0f27df85b2c24c1290d65,f98bcb1924e8c3b0e86cf418ea4bb951790d6d558e22e9981b76ef800cf54c53,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,22.901,22.9,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,7b768f670f239e9bec91a3661a853562f37850c3847bca7d85fb4db236c1e760,1b7a0443eb422764569a4d12fe7fa6c5c6a29850f406ea52d25d6599c6c2d533,86566d80a5dccef4aafa8a7f57e1b22a621cce8eba4208ae622904092551ad3a,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,47.216495,47.2,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-08T22:23:24Z,fresh,23,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,mcp-atlas,MCP-Atlas · Scale,tool-use,6a75690690167054b95ed0e7450c636e920a49ead141f93fe4a2d769413312ce,995102f2c9b208737569572df0b2bc35f034b964b2286b7b7a042977219dea38,445a8df79b0d03622ad282f4564ec2b84807d0305a2a6ccb22d06b8ee21b4718,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,87.2,87.2,exact,canonical-source,profile_86f113c90273b2ebea19246c,reported,unknown,true,adaptive,,false,2026-09-04T16:47:29Z,fresh,57,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,apex-agents,APEX-Agents,tool-use,21f0038d2e04e91f5b5421d4a429fb58b1fc13fee84afe1a61519076e81744b9,c8b53564fe22c29fb04d13466852708f9b1d857cddad41558e1389844292abd7,243df3de6e4f75baf131bbaf50a3d6cc1afb196c0993d501016330db009df235,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,59.70000000000001,59.7,exact,canonical-source,profile_0e5a1af41c411d83363a781e,high,high,true,adaptive,,true,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,arena-agent,LMArena Agent Arena,multi-turn-agent,df949d8468b4f33e73cde047633eb89fb15cf49ae2d0e70bc8be41645ed3c028,a6e21d727d920e23dd3c65fb7ebaf91daa32d98ed58186b858a8c3aaa36d4874,1e41116758bddd295832bb912d0fa8b371b05905e1dbd090820b95623d012a1c,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.13352329228313464,0.1335,exact,source-specific,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,vending-bench-2,Vending-Bench 2,multi-turn-agent,671d1042eec863892d5d5cca323b89f5c072b0666f1d6525e4e71531f3e2a779,b17184a19a086e160b569b62859d3af2247d6c09a66c582a34b7d484bef8109f,93c8c86cf424031291284c60fca053346f36cfd3a5e4fbe6d612ffe5a6219bd2,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,5421.556666666666,5421.56,exact,canonical-source,profile_6dd039f109f0800f20553ee1,reported,unknown,true,adaptive,,false,2026-09-06T19:02:23Z,fresh,55,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,79587497e46d520118b1999edfea4466960b15c244fb6f2164c93432d315f1f0,9db219377f0354a30e650c20260c160d913c190705c01b0b8d3a1ac0959f4f90,82fb140a5c92ca7db06101c5e8643c75aa52a2922550626df04eb2d37df341f5,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1763.29,1763,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,841a7fcf12d8c70bf48de536e5ff86a13f5f33bb9b2350b0cff944e6e0f86e9e,4064342f1db8988b6e75d3b6ea1ae54762ec8c59442ff94b89d217411c994a24,4e37844bb946c3b256b9b13f6de67e3dac2b3658501fd9e4610848e97e618893,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,40.0,40.0,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,osworld-2,OSWorld 2.0,computer-use-agent,8b43ae1653e6cdebd93f094dce6ad764a0733b3ba7cbcc3c93dbccf2879d68f1,833c2398e1d964938827187b07cf5459f9615f30633f4d5a5bd8e7af036d8667,11c5b8da7d1625ecf3af992e92a5a4e6d66940a641dc52fb3ea9dfa50428f88d,release_e7bb52b116c9e290ad684e60,OSWorld 2.0 · August 2026 task release,august-2026-partial,success_rate,percent,true,77.9,77.9,other-supplemental,supplemental-source,profile_c52c1ffeca7e4a640a86e739,reported,unknown,true,adaptive,,false,2026-09-01,fresh,61,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,anthropic-fable-5-1-launch,Anthropic · Claude Fable 5.1 launch evaluations,https://www.anthropic.com/claude-fable-and-mythos-5-1
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,automationbench,AutomationBench,tool-use,dedad92829a77f50ab1f77b5ace815f589a16a499d3fa805a4fff0a9e988c52c,cd36c72be2bd4128f9c213c5900fcf758ac60f5f120dc446c7738b51a3b1a1f2,084f67c3372e48b8119eeefcec7afc736f1e0d48febf04de4be378c047ba9c45,release_587f0dd1273360ff4feb0c7f,AutomationBench · Anthropic launch evaluation,provider-run,success_rate,percent,true,31.4,31.4,other-supplemental,source-specific,profile_c52c1ffeca7e4a640a86e739,reported,unknown,true,adaptive,,false,2026-09-01,fresh,61,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,anthropic-fable-5-1-launch,Anthropic · Claude Fable 5.1 launch evaluations,https://www.anthropic.com/claude-fable-and-mythos-5-1
anthropic/claude-fable-5-1,Claude Fable 5.1,Anthropic,closed,6,weirdml,WeirdML,coding,488873f6d1353f81dfce1c6a2e16f1f6697437f29ba1bcfcebddab35d03de016,10318188f10ee8ffbe05239316cb13f425c93d58001317a88a793a224eaae320,34530ceafaa84fb9bc61398ac7f4834324e260c9d0882ed93f9fe0aa2f2f6843,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,92.9,92.9,exact,canonical-source,profile_ccb7cfb699f3e49429f43ebf,max,max,true,adaptive,,false,2026-09-06T19:02:22Z,fresh,55,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,epoch-eci,Epoch Capabilities Index,capability-index,88ad41d9e4cd7b9a1d236208a70b33a8b5f5dd6c6658fc2d669073bf4462bd05,997d1a20bbe636785c5101939bce8c6a93f71fbdea6f03550995c052d467e18b,d8b3e1721f5010228416f62b6a87a36f17a3fbbc778b03e935ec33a2ea9f2ed8,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,166.07,166.07,exact,canonical-source,profile_6bc0dda6b70136f61437b9ab,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,0e5035791a4edefd00f4644390c6121f15c79b3c6ff1a377e746d1dc01984b91,d67304fdce596fd70d8b74bd6862ddee3958fccdd0711d36aff63479af822fef,32b68bebf9c0bf08b11360fe1e701a4e49e2a1f9fb41a6356122c6da033ed532,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,50.412194,50.4,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,arena-text,Arena Text,preference,41a0e2fdfec3cc955cb0985fc826e368e2e15566c6301748d204c53cec116924,fd7d92933fa6139bdc149aad6f04593b20e0674dfdcd37386448738716567729,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1520.254646597456,1520,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,f46ad3fc6765e18709cde449fa11dac57dee8736f47269d5a1b770d7a45a7598,7faa333dcfd75f920cfda03d9059a8009c372c9fd7cadf0068258c7ba3f5f62d,f0c32c6c54026bfed9142febb947cad1213b35d1212c85c136055a19154c0267,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,57.089898,57.1,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,simpleqa-verified,SimpleQA Verified,factuality,e42457ab3eae666fcdd578ee756ce552be605c7500ae7ba8fe889c6da5aef70c,4b42dbf1a77298fbf58d5d6c19b346674cbf56aa26869db298f4a83da55b2d8c,51e663b336389d574c8880e0231f6ce4803afecdfd2e0be16c2db60169d9a747,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,76.0,76.0,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-13T17:15:14.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,gpqa-diamond,GPQA Diamond,science,942d1e3e3cb86d53e71db4236dfac74c8028b79ad9e6c202ac764897c04b494c,3605e4be7b17d1ca02e7d9eec73c3f40963650e9b84f39f9b2555b397afd790f,b9f9437fcfbe7d191008e292cea0e920ad99813aacb47cc16787fd687f0fa77a,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,96.40151515151516,96.4,other-supplemental,supplemental-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-13T17:15:08.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,49dab35201d4145d47c970c96ac0a3a408f86a49cc0b364d627d6a87090bf0fb,51be0c0cb0e8436bf3a55c6ac42e70978d45e15dc14917a1dfcbb00918c7b6b2,e2199c913d23a0fd6c1053b929330bb596fec8542b3d5e7aeae4f789d87fabe5,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,100.0,100.0,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-13T17:14:57.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,frontiermath,FrontierMath T1–3,mathematics,ef0117d25262ddadbd4f13c7dc345baa406ddf6b8e0338a2fb77234009d3c94f,acbf7c56fc4ef7463c9ecd6b3c0b9769f600b7cf89596c0dd682e51b8edf7612,24fa26dc27b6119b2f23dccb7c7da348884c4145d0cc9d02f4674ad281476673,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,85.96491228070175,86.0,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-13T17:14:57.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,arc-agi-2,ARC-AGI-2,abstraction,cb3959d6826dbb0139defbb18ecc0e468ae659e5b0c231c9681a218ddd31a2ad,eac6346987f2ee5621d0f633f76f941222729ed4b95326f40c5e66fb3168aa69,ca829b83ba736df84b0d42690b8fdf5d335f17160a6c1fb4034209c38c9c11a2,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,91.25,91,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-15T18:19:38Z,fresh,16,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,critpt,CritPt,science,5212a92697c420addc51b40bbeace36f572558146ad36fa1e9afe3a7e7562311,4e11548fda6a4fc2b37263feb672cd9366bac66dbf131c1656ba33c986d09bc3,3bbd3ee9062c5777efaa7a20a72a8fd62079ff597fe8b7690f28dfea0f985e08,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,27.142857142857103,27.1,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-08T22:23:13Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,chess-puzzles,Chess Puzzles,games,d462a1afbe9e43e39671aa888704b349ffc11b9e1598e074fc2b7f22ba57638d,0c2d2f2514b4a7b0d34db417c75761739061e9c57d017b3509df8a393c2ebc4e,593d4ac84a5e773c161de45ad4c3f8fe1294c75342486875d1ac47a75c9539b7,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,59.0,59.0,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-13T17:15:18.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,arc-agi-1,ARC-AGI-1,abstraction,693ba0b7a4ec5782d44524e7f27946eef47aa3516d119bdb27cf4d9b61ffbaa6,fdc8509057cae5509564e16d8b33881c13404624f47d19094b9f2c500724b739,c8fade6d8de8a2ee6d3e84a3272be7ee289917fb45c5a589964003c61ffdabad,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,98.5,98.5,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-19T22:24:17Z,fresh,12,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,deepswe,DeepSWE,coding-agent,bc60da63a84afa46bb438fcc68cc9171d9e1d5c0b1c0dfcee13ee7a59cab4da1,7ca21514ef0f05048c3245fae40a64c28a0c94f3cda4615d293ee3ff8c53e80e,7cc192befcef15af5483c5f2ef32156eeb6489f75d15f93ac8576d1001e08a8f,release_12e8a62d68d933c29dcdf020,DeepSWE 1.1 vendor run,vendor-model-card-run,resolved_rate,percent,true,77.9,77.9,other-supplemental,supplemental-source,profile_31e1c3f95866a70589539280,reported,unknown,,unknown,,false,2026-10-13,fresh,19,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,google-gemini-4-argon-card,Google DeepMind · Gemini 4 Argon Model Card,https://deepmind.google/models/model-cards/gemini-4-argon/
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,2ffc7afe59843f10a158700bb0b58b7850fb3ae22682af6142fb8ea7a27a222f,4d1cc362f62d471c7bc31bf837d4c43aa402c5a0244e9c4ff900c8560d0ed3c5,7e33cb05ef7be2a63667dbb40f465f8d7f3f5128d8f1d0528fc5de2bbf642ad5,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,99.22,99.2,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-14T16:34:00Z,fresh,17,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,cursorbench,CursorBench,coding-agent,da8e76103e76194f94010b0e117b57daa1d955f95a53f9fe24e354be24e3d76e,ade7739c3063f5ae6dac1b0f0586f1f2cae944946a0cf6429b6d125f597316b2,b17dc5812f51c58f6597cf0dabe7d7888f17f0f3823affe5bbf0ecb38af450eb,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,52.900000000000006,52.9,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-14T07:48:11Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,frontiercode,FrontierCode 1.1 · Main,coding-agent,ad5524d876fdeefcb4d869cc3fc792f9ef20011313690adb1943a4dcb09e8355,0384539aac00dc5622f9bd67e9d4c171e80c0157a268622663b2b9d65b1a1ca6,d9420a0001ac4d52bb3a9ab2abb497f88338950ec67a761eef9db1fc229dbdec,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,47.55,47.5,exact,canonical-source,profile_9d3f500136d0b273203f54c0,medium,medium,true,fixed_effort,,false,2026-10-28T21:25:01Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,ioi,IOI,coding,d73442638043fbc652f1818fed6700cf4bf7f8a814e2201bb51851fca6b61b62,19040fb900226cedd7cb49ab89f3aeb284be0e23653ea0d9bf4ae662b9ab5ffb,3e09044ff0dfe5b65bf6bf00c43446298ac1023e77dcc215504a78628545d45a,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,100.0,100.0,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-21T19:24:33Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,scicode,SciCode,coding,f9fbd84ed17d8d654fb21447bda0e04683ff0d7cea83caa647d29f862386df9c,f85b9c2ac70884de215e722b7bbb4c7eb2a8b2fe1f8521ba5be7ddfe0d0b7a6f,b9a690d5875d207c07df50139a8b4350a3e2296f506e3210e323dc88107e3569,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,61.8055555555556,61.8,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,ale-bench,ALE-Bench,coding,a1bd32d8fd1dfd5ec8201c3dce67f0a940e792f1d0b249b2f7a1f7fb0cf8bed8,9902f81c60546cda59c6a86c319941e076b6c92d7b5cb6c5f9c4690b89400cb9,9cde006f9bf8041e1911ba33fa535723ec88f1f80ab0b2eec89b2bbdac8b38b3,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,2100.1,2100.10,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-19T22:24:18Z,fresh,12,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,vibe-code-bench,Vibe Code Bench,coding-agent,31545f5414896bdf0fd5113fb5cbccccd6cb392872c0e9ad858d4909fd4a74ad,a1aed8b102f6b6682b8ab5ba3143e24123376d979f9b7c78414549a3a25e8a40,c6a98a5c0d2020c6bd69fddbd129a50d8adbff22756d0f87bc89f0fad9ed22ed,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,91.906,91.9,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,webdev-arena,Text Arena · Coding,coding,1cd00b8b3a6f4ce5826a128da224cbe100d784ac842bd3f8b34c839a778e1a26,b43706dee35fdf1c938e16bc42b49d2f03b5359cda4bbd2b0517a1fcd4cd5910,bc0b5ec665b2e642520063516135b8702fd7c46e36ccce33136fd3adffad9533,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1679.2118291253184,1679,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,mirrorcode,MirrorCode,coding-agent,634003393a66efc3036f689328c9f5983a5be6f8c4795194a9c8bf93de93576e,ab562a05bd21f8dd1c02642fcf256737408112a28d74a1635ac5171a28f566b2,f3c792e2e350897524ce130389403b01078f6bc176705d2f90c8857cb9417369,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,58.88888888888889,58.9,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-14T10:00:34.000Z,fresh,17,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,programbench,ProgramBench,coding-agent,95b9019a140ca8df41390cae58efe167f4ffbd646a25f1864c3e53d855a30f17,5f3594a351fa09eb1e52471ba8e33cadc2c6e865b7ea10c07a6d65669b3f5e41,186e0b0f922ce72c966821b283b21dccc813c1552b6115d47b78cbaed52b12ae,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,2.5,2.5,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,weirdml-v3,WeirdML v3,coding-agent,2cc79e5c7d9b4894b52f34ae8d3538bd8998ebd83127663477b4d2b287096676,288d450f372d25a72872863551edd96b74f005da0e89d99069c9b197d4985e65,a8f04c285e30e485e3941bcbfa28c41fba5dc2a6093d65e0148d7a16ecfebcda,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,20.096489621212122,20.1,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-28T21:25:05Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,d5d93693bc9a72915d0d01623878672da4f686038dfa464039ed069477d640a3,dc5a81d5b1899a7efcfb990d69249f72a2454f4b278b8b7433052c5ef83b181d,bdb80c3a4768f033af8b48b02c804e607ea4713041d354a06757cd83750ee2df,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,57.678,57.7,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-26T22:05:28Z,fresh,5,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,e9670b81969a9f02ee242ee4cadd7fe213752919c846014355f97abfbd224898,034b4f00d4108f17940502ad41627b93d93f192396f4d3fbedd01c2065a5ba45,60d2549c79da03f7d575e554a4a1118890196bdbfedc204b1e290939fd5c773d,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,57.576,57.6,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,sre-bench,SRE Bench,terminal-agent,cdb5c28fe48242f3f111f8441f30f7883fbcf33c3b51a8e2cb53df2cfa6279f8,75a5741cd5673547462e450e683be6b30448184dbf61230a54ea6493a9738396,88fe646d7e267129dd890d44eee457ee14883b807f3a02c4133704ea122d099a,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,44.275,44.3,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,apex-agents,APEX-Agents,tool-use,b0e7caf390ce08e63ad5f7c9e3a99ac08bed3a65b5ae783dc8084cc2b2e1d280,ae5b5d55100706d6596353edb8e464f67b705f90f38f06fb007a9c2451c78a26,c51ea81fb0bd43daf734c456feb0a0a2ece36e84d111c2fb28fbd589ab59b5b9,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,82.4,82.4,exact,canonical-source,profile_88d8753549159481a7ae44c6,reported,unknown,,unknown,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,arena-agent,LMArena Agent Arena,multi-turn-agent,28e75ef62c1a520dff887c8d56d6d564fbbe1be1b7026339954bb6cc30e58d7a,750dc66c5385fe45d26b14c20bc20e7aaec0e715101395190df4560c43dd6cd1,54137ebd27d7f2d145e32a28df61d4963b65f5642674011854beba33cdb46402,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.10727188431308599,0.1073,exact,source-specific,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,vending-bench-2,Vending-Bench 2,multi-turn-agent,05118678e5933c7560dbd3f43e95e16e9f9fce157c06d0bae3827473f67109a5,9314bc5673ec5cc899e801ce14e1c6f5d29dc3abafa9262fbd45e634153c17b7,b0407e9e0a27cd895b67f5534f82db3be2e20b3c723f2d5c4729ff209af96a23,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,13718.158333333333,13718.16,exact,canonical-source,profile_0db4a2e4aa983274af2ea61e,reported,unknown,,unknown,,false,2026-10-08T22:23:14Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,44b8be8cf89a254003d2453058b2dc484c119fc8cad372d78732e64325ea0ea9,e22fe13392e187fd01d43b76cc365cb3d9d29394d3b11654fbb9fe8886339651,71c04dbf23b2eb31f190eb5372ff4c90a359b3c945c1ccd5c8e7271299b530ee,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1630.66,1631,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,8e3877e37cdc012ca8f84a0801568f74f289705df3c06eb73a11bfbfee50a056,295e099530979fca1ef5832e566b3a3a646d5d6629fa27bad62c7cd9d1ec960c,bd9432e2786811a0d3deb21d1c36bb11673067dc745543e04702f727d19c77b1,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,44.286,44.3,exact,canonical-source,profile_bdeb2b4867ec518e7c0919e0,high,high,true,fixed_effort,,false,2026-10-26T22:05:28Z,fresh,5,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
google/gemini-4-argon,Gemini 4 Argon,Google,closed,7,automationbench,AutomationBench,tool-use,c0d5e0b1871a4718bb68cfe3e9991f4a36b3cb928b464b7e1e58f383e62abb7a,366ec9ae87fb412650fbf5aba6d03ce8d38332d9ddb097826e5b70d73b9630b0,96853ff1b468d93051f396e775a1a7d9a0cf86ccba3aea71068ba91fcf4b5a3e,release_0e98d905ec617087da8bc4a5,AutomationBench · official leaderboard private set,official-leaderboard-private-set,success_rate,percent,true,51.3,51.3,other-supplemental,source-specific,profile_31e1c3f95866a70589539280,reported,unknown,,unknown,,false,2026-10-13,fresh,19,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,google-gemini-4-argon-card,Google DeepMind · Gemini 4 Argon Model Card,https://deepmind.google/models/model-cards/gemini-4-argon/
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,epoch-eci,Epoch Capabilities Index,capability-index,90481d6c6f72a551d6f55a3e60f559afbc605c98135f512519f8b76d7178c4e4,7069c2c299c5bcbefcb19d00873c1c3ac11c626a4e5c6ac33f06cb767b227bb7,ea46c241710ed2cddb9f9e7d74b8b827b77077865d2796ecac5efbda0fce7bdc,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,164.47,164.47,exact,canonical-source,profile_564261edace44339c4aa1271,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,arena-text,Arena Text,preference,2038258535c68c2452d8f04585730b2d292f048610b6e86cd9fb9b1bc7a5f52d,17fec020e26c4795c461221a4426cebff1c295ab8f879dffed5a837aad7048b4,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1490.5040398730391,1491,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,livebench,LiveBench,general,5804e2d7e5ec1a1f5599551d3400bceb9897dbd45bf7efba5732ba11ce522fd4,27c26a3ef66768a6d5fd32c1259722b1837eb4b670b2ae6f860c2a51bab7756e,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,80.08497619047618,80.08,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,5b1803aa53ae37b78e775ac27f61a658a307f6e77b3e139dc08d064e7ad296fd,5a69b22b38dbddbd17a3674fb69aad64338d3e3f2220461284fccd7ce033265d,1d000b8ca17f30a20519c0fbea4c0033b69f0c618060fec604365eddb311a0c7,release_c1afb6c54b8ceec8bc1f121c,HLE original full multimodal,full-multimodal-2500-no-tools,accuracy,percent,true,56.3,56.3,other-supplemental,supplemental-source,profile_54c2373720fce61d7d0d3fab,max,max,true,adaptive,,false,2026-07-24,fresh,100,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,anthropic-opus-5-hle,Anthropic · Claude Opus 5 System Card,https://www-cdn.anthropic.com/ceaf5c7ff2783855203fde8208ec311252dced5b/Claude%20Opus%205%20System%20Card.pdf
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,b17f9562b14fe8d6b71a79b35889d501511208fb1b24913122d7b40ed715daba,b3ae8bea66dc042b2a7b0e1933fa4b35b469026c406e585cd0477dcfd9e94013,1e81256caf684d8eec577c6a8080a1e254f3004233c6bd143c67daeccd176ec4,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,54.865616,54.9,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,simpleqa-verified,SimpleQA Verified,factuality,d17c208ebddb7f596810111b5df5614d261a6fa6f9fc1924009880646c74e412,7077b7de72bc989e8953fd3023e4bc62813d19c92ea57d361d08f44ac39b6895,fbb17761ac6cb9182bdfa96233250087a44472d8d76f545b1bcbafe7cf828e8c,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,59.9,59.9,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-08-10T23:44:04.000Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,gpqa-diamond,GPQA Diamond,science,19640dfe822bc6b29d77c4a1e1498c42b7f2a7248ebffe113710074a34c0c53c,612a2d94e1582732f9923f5f37c7f002d6fb2dd4ed1f56f7a9c20b9254c3c992,d3777f133bba23744b1423cf6a2451a2cf11a061e1b5c5ee1e57940741baeb43,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,93.434,93.43,exact,canonical-source,profile_10853c414d0b2e6927badce7,reported,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,mmlu-pro,MMLU-Pro,knowledge,7f7e209eb4ee1611eb628f96a63fc4e17f6c92bcd1b922f598d9ca06c8ed9cfe,eded7438ac1d156b0b01d6df0206d538d566ad8a51d27cd40bec4ccbf527924c,b9b1fe4354da8dd8e5959b0fbd0f6af45b8632e77afe9df327d774f9d2dc616c,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,91.591,91.6,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-07-23,fresh,101,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,36a7cfd23cdf5d76a4b852b810ea6a98fe77ad4288676e299c3b0caca9cea503,8f7ee8e98117813caca0c1c92a8b46158defa354343f0c8937fc98797c83d6be,a9d9b4fc68968f9abe8cf470be2c363f3e52f5d85ac66dac5e30fc8c98d92115,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,98.88888888888889,98.9,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-07-24T18:40:39.000Z,fresh,99,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,frontiermath,FrontierMath T1–3,mathematics,f5d71d252043fbe2ab87da47308a2e53287b7f1add4da2ddb0f89dfdcb0ac622,16a7ef48859029a41795bcebb37f84562ca06a1dbd75992ca3ad3b2818efe93a,e8950252a5c51330728bc6367a249c0cc704fac117b382f6fe9b1d068b0f8aee,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,85.6140350877193,85.6,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-07-24T19:54:20.000Z,fresh,99,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,arc-agi-2,ARC-AGI-2,abstraction,a3c4daafd9411c35884f6889981b8fe8755d0c681968a77fe06819e35f34f0db,b47713aa60f93c953c7dc3ad93a307a14f1aed2f7085cd87da6dd47d7137ad70,e4375a17736c8f613f4c9ef2a1a2c24cc839e2fc69691eeec35e9a9796c2841b,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,90.42,90,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-09-02T13:33:02Z,fresh,59,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,frontiermath-tier-4,FrontierMath T4,mathematics,7b52f695de6f48cf0ff1ed1a8bbadf39cc4a03e211961ac94802b81e1f396629,eb3228eacd51a5fa15ae45076fb9c0053909413ee24229f0854630f56232a451,e3cf9103efd2521f7be4eacfc07310d73b84b17da7d89c4f1b7663c6396f3cf7,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,73.17073170731707,73.2,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-07-24T19:54:20.000Z,fresh,99,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,mmmu-pro,MMMU-Pro,multimodal,8d036e718cc8bc18dffb44cb8f7663d64e981d336244effaf5c7ea11f7f0afd2,a8b84451b09064fd9fba5d3668fc3116a2854c147dfc6112e82e6921c70ab6de,f37d2f335aec4e908cb99d6013875ce9e43ebbf3f1114afafbf6f6b07913d035,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,89.884,89.9,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-07-23,fresh,101,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,critpt,CritPt,science,e0db2cdd2701de0824189968923f0c4060f65cab29a42108258f6d2653bb18d6,9e9d2e6ac44cf398b5314d62ccf82b672e63ed6b43733d30bcf50626bb2c4d65,09b05cc465e646d9e64a872d732862c4548a4d4ca19620c3b74478c312069c54,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,29.1428571428571,29.1,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,simplebench,SimpleBench,reasoning,f4a1394950dec513d8d27cd96bfa2813780e293c7daf53404107339f6f8e8f8b,39d0c5fa268349b1c52553059128bf3bcc8b0614f34b221b365cae78e2ac5c75,45dbeb835a69df9d86f682a1b1ef7127e521f78f7e45acad94d8565d9d92bf83,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,80.6,80.6,exact,canonical-source,profile_e69018c538773cbfb94889c7,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,chess-puzzles,Chess Puzzles,games,8a82c900f324fe3e3d4bb60becfb564a9af722a4a61cb2818f66d6493e66ce42,81064f5fa7c3f56619fe95c6b4c34f0e8c8d48f5ac72e18188cf9f1482e9ff2d,fbfbf1dcd93a5a55dbd3fcefc26accb3ce1f04972e0f558268e7da5ddb3fd9d3,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,43.0,43.0,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-10-13T13:19:00.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,arc-agi-1,ARC-AGI-1,abstraction,71e45c35a7dc614cdca26eac505e5527c9d928cb91b3565378fab2fa19314191,ff084937372d7fdb731ca84f03c8a24d133f8ce1f46d291106260614bfd375c8,b3ac1059018444402272d7dfd2f3ca5040faf0501c3b838bc2934dab3b7ccd96,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,97.5,97.5,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,swe-bench-verified,SWE-bench Verified,coding-agent,ea7569efe1075f992145a9a17cfad4a510fb290a3f9bbfdd79dff3d4a04f246a,fb5197d4d2512514d1b70616bb46b70e2b43f2902b8e3ed454150e47fd584ac0,99d9805a5246664aab20ecc8561bb9de0b7d37ae92eec218f614c2601eba72f0,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,97.0,97.0,exact,canonical-source,profile_37b9c147fcce0c9004cd857f,reported,unknown,,unknown,,false,2026-07-23,fresh,101,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,swe-rebench-v2,SWE-rebench v2,coding-agent,18b0518de7854e79588f71d6580b3d07a186ebee8c2c382620e227ddfdb5710f,7cabdd002693a3078961d470f44b203c8c256b056487041f92cdffdd3829242b,7486ea3b687109a351af3f6f316d3157c470fc9b0cf20882bf67adcc9119d146,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,63.42342342342343,63.42,other-supplemental,supplemental-prior-window,profile_39a30289a15f1db029022e6c,high,high,true,fixed_effort,,true,2026-08-30T11:02:42Z,fresh,62,0,A+,0,A+,7,C,potential-contamination,6,potential-contamination,swe-rebench,SWE-rebench,https://swe-rebench.com/
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,deepswe,DeepSWE,coding-agent,ecac36582e53fab2d8740ec8a0e7393c1d7c6f47e82865750f540e14d66a9c15,f3f86ae4fbc2ff3e6d57e2bd3839536265a211be3cadcb7241268bd4a09778e7,ace259fd68d7198838f969ae1b26d71e836fb71ca059c1b781a51b215bf35605,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,73.64864864864865,73.6,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,31ba3711152e42aa6042b46b19407c8d85d53a335e3bbf9ab74bac49f7fc4f3c,1f4006a3883f1f54a597ed5de60b46f704c559c233edc825534faed604620fc9,af40045797469b477942e9fc33dc00f5583301ed8c5900a0f3ef8198d6d63665,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,99.4,99.4,exact,canonical-source,profile_1f510fcb6cca3743ad871cce,xhigh,xhigh,true,fixed_effort,,false,2026-10-07T01:44:58Z,fresh,25,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,cursorbench,CursorBench,coding-agent,96e0288c9311432827165d9a355993f8add87f38f14584ca23efeb076d7194b8,73e716b8e1ca93c4308f6d60f2b5fef372800f5e9583d888959b2cce3fd1b01f,c570eefb08611683c6cf09a39aac7758ac0714c34f4ecb5ca51d5e4d29150a83,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,46.6,46.6,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,frontiercode,FrontierCode 1.1 · Main,coding-agent,62f5a0cd4b228cb5aedc46843d62ebb6585683def5a4bd7add55e8b06ea01879,83fd3f0f1f59575a2ebbd4de32bb8d79a94bd80ef10082f1028e99e2194864b7,e0feb047845803ef3ff53270763eb37c74610e5ae13172e2119df2cf213d49b3,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,53.38,53.4,exact,canonical-source,profile_7755c5bbeb7e47ccfd5ac6c4,medium,medium,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,ioi,IOI,coding,f5e20de2b8c2c1608413991238e73b085ef75a671e81a636296ff00de3b1ff5d,4291d8fd245bf4e8c82692959c4fe208061771c89ec2165c6ec7938ec2b6a701,7a4f80d8916be3d88fa02c0f398788a026eaf7a1396b4e20784cc5a5cbe5c798,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,84.333,84.3,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,livecodebench,LiveCodeBench,coding,e6bf68c5c7f139e05a6aaacf58e0c7090e6f1290e11bb2dba8b7ce4740acec58,ced82a3135dd3360d0d0a52e42358ec222ffd48305627f98db910d7babaa0a0d,5ff8fdb52cc4957ca9d9313f813773f22df0ffb3dee80a5f1369ac967b689aee,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,89.033,89.0,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-07-23,fresh,101,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,scicode,SciCode,coding,49dd9702e4e69984ec4ee37a5c743c3db7b22b84ce8806efec896cff99d0ebe1,35e0121a42bd54fa26aca412698bd380f4c2fb0f2f5f3e41f9bb26ddda9a6982,cd9e7560d85a0322d3b34a586a289f41e6311f90632278fdc833038056ffc518,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,56.3657407407407,56.4,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,ale-bench,ALE-Bench,coding,56aab26add6a820db189af4a3f80f85dbb66f2f35a163a98239f5e921cd45adb,adba2ca9d3cf1a4502e9cf9104ba6783ea646a7fb2ac51b49e1d1c8f1bde6138,f5d0d49ca2d1395a9572a75707f8f175364d159218fff2ad97b2fb9f4823395f,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,2164.57,2164.57,exact,canonical-source,profile_39a30289a15f1db029022e6c,high,high,true,fixed_effort,,true,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,vibe-code-bench,Vibe Code Bench,coding-agent,98c4ddeb8d354c4b8df09717805a7c7f62ff4e34e0d64fe9b99f124c32fe0371,8e44f161ac4d12d30044e6f82edab9d84bf5ac86744d8d17ee99fd01253f9a28,ecbf2c6f455534b1098adf00c37ce0bf5466f0c96902095e7437cec06808394f,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,88.404,88.4,exact,canonical-source,profile_ed92784a3a69757d5a7d96f7,reported,unknown,,unknown,,false,2026-07-23,fresh,101,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,webdev-arena,Text Arena · Coding,coding,45206f0e5175578cbd241821c362fa5b02101f8206cffbfd45e99128e9335f22,622696cab4e4b65cddc0ede9bad78e63dd28402bd2c72e5d875752145923b64a,4a21fa8bfa35e740283ff1fb138e267eefe15e1b1075757f3af9478970704d8a,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1690.5239169555273,1691,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,mirrorcode,MirrorCode,coding-agent,59f89fb10cb1ec882a01229237eef30e31e47d86f7ca49ffe4ebea045c2c0d37,1a70bbad68e6959182c78917cbe70e067514d8ea0fc2dcb1d3ef91cc9055d749,331a82c06468adc747ff58df364d74ffc2f7fbe2b5d6ec6e706cdcc5f2336ecb,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,87.77777777777777,87.8,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-10-20T20:32:21.000Z,fresh,11,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,programbench,ProgramBench,coding-agent,10d2a000a1b6795c95a7e16f72e6cecc66774b653e343a920b45f14b5d548ca6,727772146520f09a3280c5bc3e87ffe64b045116a913615e528a781a856d3373,86f22b6a3209094afa1cd49f8c0aa87ffa6a73fb2c6b1d5590dff894c33e5945,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,3.0,3.0,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,weirdml-v3,WeirdML v3,coding-agent,c9d7eb78a63d9c3c28e15e5c2e673f3e57732b390ec9455787de4d469e35d8c3,3866d08f40f87fcdaf0ea2eddcf2735d89114be79a96808b099c591e5d092a9e,913ef5fa9976cef286ca0eca6e352de5bf70b90e94e06797c7d7e94c9957ceca,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,19.01117537878788,19.0,exact,canonical-source,profile_1f510fcb6cca3743ad871cce,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,1c61f901c2556734dc534377074c34016966775866c9e9f9b36a1db4d00915c7,847ffe07ce22b665b4d6506e02dbfdb23d65eb667dca9d568babfd9da2a1ccab,4d15304576a02115cc358aa707c120be20d7564e2737346c0d7a69b2ab2bcfe9,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,84.644,84.6,exact,canonical-source,profile_39a30289a15f1db029022e6c,high,high,true,fixed_effort,,true,2026-07-23,fresh,101,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,f36d2528199738ac88394386d15c53b26eab6ccf73bb4250719d3f5d25a282ed,a63ae0c93439c5efe64c6b0cc39942f10155da922c4b5998da83a4c051beadf6,5988374bebca1d4a685e6a5a7ce86a373896d268560c8948d3c339b7c600152e,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,53.535,53.5,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,sre-bench,SRE Bench,terminal-agent,a2c28f2560197339892088092179f16c6af6ece271dfd9006affece86c05e4c1,07dde57d1f5502c2d4a5e00c93359a47d8e8451ea3e756bc7d8b6028f34709c7,aa4ea5f3bc4684b3f5e30a236b21d41ffcc0166fa5f2639024953fb121e050b3,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,12.214,12.2,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,1ac3ac67ac28d441b8885d34ddcce4c7a7cedf782c4a54a4063a1be27556438c,29519ff8c1a2f1c4b11ed1a5ba1d6a524f05d7ecd4f3484ccaa730503d9b9391,154f156c90ea702c55f0088b80f24f0bc7774ad5213995666b58947943052ec6,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,42.061856,42.1,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,mcp-atlas,MCP-Atlas · Scale,tool-use,7b53a9b230f8c418e679e980b9b4b41ed0f3da5d56f859877922d319c114dded,95b7008630b7a8537aee4e980809df66734d57a6fdf0903c50007244452bc9de,15fa6af638e60bbbeec94d7790a5e8e4a0ad15b1efa6f42264a3dea944b1687c,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,85.8,85.8,exact,canonical-source,profile_1f510fcb6cca3743ad871cce,xhigh,xhigh,true,fixed_effort,,false,2026-08-03T15:52:23Z,fresh,89,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,apex-agents,APEX-Agents,tool-use,ea83dfaa4fd08de872a7c9d6716e144c9a994337410adb245aa8303362c52a7d,ac85192e0846a1660e62bd5f98a737ed6d837b94fe380b6e62a87790fcd77e61,5515c94e4576ce43c29b263701e65432246a210f2bfbb399e75c317c22bcaf9c,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,65.8,65.8,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,arena-agent,LMArena Agent Arena,multi-turn-agent,f8d674c296de7f6e33cb8d87092a5d8cea64ab6769f19daa52ca7232cdcf97fd,99ecaae24f05bbfcb012dd0a9e874902f02acf13843689d9896b88bd448b1534,702a85a2ffcc161882fcd9663e2ec294bb4de0d98e9d34b97d302e6aa47cc730,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.07565942758202185,0.0757,exact,source-specific,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,vending-bench-2,Vending-Bench 2,multi-turn-agent,26d2fe2921b04a5858580a30ec05ed0c17cdae5c2a91613da3c9d11364ea2258,0de88fef3ff8d1446df66f86740b5722d8bffcf442e6f1b6b0e3e80220347ad2,ef5249d0c3264aa5f421a764813869371b4d544ebde5029afcbb4ab79490bf64,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,11181.87333333333,11181.87,exact,canonical-source,profile_1e112a485e54e7367e8e34df,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,bbe63543467b5bf2f01362f25c4c373ab5b5dd11f36cd27f17f671d2146cb17d,08ea9618f37989828ff2f5b43af759b08916e65307ed88febd4ae46db152f049,641d7cac73f40b65ebc19088af68fa9f03b1c1d6b2f7cc17e618b857d6173af1,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1729.9,1730,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,terminal-bench-3,Terminal-Bench 3.0,terminal-agent,c3d9b14837df09d99fe6907ac42c1da03a61d742061c11370e38024406950436,6871dca2fd8dd990bd67c2b29d8582c79618b751b4aaf7858a370ab31ad98dca,5131b7fcd81347976de3f3afe9b5af9ea833e46f36bafd8cdc477f8e699e3974,release_ceb36f25882d8cea2c9d084a,3.0 · b936386b-4afd-4ad7-aa7f-6b52bcab9853,official-74-task-leaderboard,accuracy,percent,true,42.7,42.70,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-08-12T22:08:44.699364+00:00,fresh,80,0,A+,0,A+,0,A+,official-leaderboard,4,official-leaderboard,terminal-bench-3,Terminal-Bench 3.0,https://ofhuhcpkvzjlejydnvyd.supabase.co/functions/v1/leaderboard-read
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,e16963f2d47b8bf2f6bd54d9e4db52a7aed69197dc0099aeb29480724f918cff,6f36e9bc4765b8d572ecffa89ece1e05170a8d014260409c093a384c4695da58,44ff308b1784fdc20491c360809c9b64e80834fbd42894c4ccc08a9a65e97dc1,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,27.143,27.1,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,osworld-2,OSWorld 2.0,computer-use-agent,e0fd9a210c3ce382239520db26581f54c924750db93b25bb31dfd17c908febfe,a17b69654848b3cc40b8eb8a1c36997abdb6f461a1385fb93690f55c0a926c9e,7b8b7a1ab940a3e15ada25e67e664fc8cc3cd93e799a1174dfb120fba783b039,release_15c5525fe39f33bc181dfb69,OSWorld 2.0 current leaderboard,binary-accuracy-500-step-budget,binary_accuracy,percent,true,31.430000000000003,31.4,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-09-02T21:27:40Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-osworld-2,Epoch AI mirror · OSWorld 2.0,https://epoch.ai/benchmarks/osworld-2
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,c177741c149a615691d2676f1bfdda13b33f6752f927c3bdc8df2418e40c46a4,5c96e3eed570ee8ad72d91901a71441b7b130ce45bd307ad0595748d884f288d,b506c9b684e54c4990e42c89c90ca0191030f4ede4e7efce735f197e2f9ed3e8,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,48.71134020618557,48.7,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-08-03,fresh,90,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
anthropic/claude-opus-5,Claude Opus 5,Anthropic,closed,8,weirdml,WeirdML,coding,5ce07bd36fc8b21acfa77c083d8e9c0df894e8d4a927123480d0c05ddab32bff,89b627bc672914d2a2f7240a5cb783d18ee0d5d4ad9dfac82abbbd8c5d416c3b,329bc9c59b122dd43f418ca50148d734057d92ea5b39e2b294aa783aeba0d7f2,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,91.78,91.8,exact,canonical-source,profile_892e19ae915f3f7411f86632,max,max,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,epoch-eci,Epoch Capabilities Index,capability-index,b803aee47f91d296acc266d790fb8fcebcc5a64d10d133b8075bfb4023e0bfb8,c210bf30ad121911e38adcda61834d801da66679b498bbfa7fe8d56309c1427d,bdf24cef4a55553ba7ad44003445881c756bdc2d3eb4950c44b500d538ee1b18,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,163.78,163.78,exact,canonical-source,profile_adcf9c2d99684658e4bd5117,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,arena-text,Arena Text,preference,a5456a9a2ebd58c6ac8fed2e4ef2fd909e145f0829c1f47ba6359288eb11906a,2c01d7c091206de4df601825a473716d7c70e5f4f03b71a4aab92b66f1abd99d,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1458.3170296467606,1458,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,livebench,LiveBench,general,d9525f8df900b6199ab887afa0fd236512c462e2f492a50e8d65cebdc7400124,0cabeff69b5c782657de8ee8dc87cce2e395912adfe956159cbdf96417f739cf,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,79.24545238095239,79.25,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-09-30T18:53:19Z,fresh,31,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,d2799c879f69385b034789e5c680fc433a327b468371470fd4ede7756429e31a,c70db6e58813023b7a184601283a3c6e50b0f18ee3235ccb780c2193945ef9f5,6dc7bf79d149cc8a5b818bcce561f84eb258ebfac9bbed142a05f737636ef2f2,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,47.914736,47.9,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,simpleqa-verified,SimpleQA Verified,factuality,97b2de31efd3dbfd86fd5466c06f6aa4f32936b5749510a7d1626a59e9511583,8d8eb39ca3be7cc3977ee9c5d09f74ba263e67c22dc966591d5b1cd58c8fbe84,418b64e983ba024c19a2a130b1474da08ea4b1b41b7e78d42b4539f491bb8e98,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,60.699999999999996,60.7,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-09-22T19:59:24.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,gpqa-diamond,GPQA Diamond,science,84225bde21e78759a5d942d1b662bfb0b4024731a221088036085adb3d26d764,ca7151100f115dfc1763d200ec929f006735e6281bbe33f6d1e241e623845e11,a810ed65ab93af76f7fa38055d2734f05182535f56d8816c76257cbeeceff2e3,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,94.31818181818183,94.3,other-supplemental,supplemental-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-09-22T19:59:24.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,2618a8d6a667d4af1ceb34799b11b013e2386a9f103178672e63b282cc65644c,537bcc9dd80885b8a8643cac6e7009c4ee0b827cafa53eb89e53095580f2ac18,c54492e9d4068e094b3ae279760a6322c0ae91c104cd7424160602b3e18c3af1,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,100.0,100.0,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-09-22T19:59:24.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,frontiermath,FrontierMath T1–3,mathematics,c54c7dee8d2a1e235d78636c59d44e236ac2efb542d9574c9af04e9343d298a1,ff44365e6747a66d27f01de30f303fd60b7b716da22050b7720959e35063da65,c79b1fe7255a430d7782a3c82799bd0574be029d6b4d8fd89c716a787ac624d3,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,89.82456140350877,89.8,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-09-22T19:59:24.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,arc-agi-2,ARC-AGI-2,abstraction,9fdf04a679bd41cc2f6b486b0adfcb1b0b835b5430f6efe3d5d2106421a700cb,39140efb96e55fca94135b09d5e0cf7c95d48ca39d1107f4122f74caed23f42c,b4e761fc2548ad008906036cdcdeed912c2582cf7697a517cd71e289a02236d0,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,89.58333333333334,90,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-09-30T18:53:20Z,fresh,31,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,frontiermath-tier-4,FrontierMath T4,mathematics,2d5892a79155363a655cb6a2c0e1ff7018efe96fcc22fc338d610bb59eedcd2a,a3ef1c5512bf41b161fc2fad979e0b618817c9553107376550c3c8be58896863,5eb0d83a930ef1849d72bf9941be0f428443b42ac0a1368b3d6eb934ecde6f2b,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,90.0,90.0,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-09-22T19:59:24.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,mmmu-pro,MMMU-Pro,multimodal,e65c1ea0eba274f01cd8a220861f7480b1e51154ddc4d5090768ee6d14d9fc47,dd89003585964b02290da255b09e95414f7a8b6b844a49773c7f35ee9a65950c,79d7e035e0b2097f236b40d1ac9970abf0f6442601c9cce7248e27665d4b93cd,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,82.947977,82.9,other-supplemental,supplemental-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-01T09:31:57Z,fresh,30,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,critpt,CritPt,science,95882c395dc0b0ecdc7050cf168efb0dfcd46409fa04499f73f8e7ceaf671e65,0d7527306ef3428b7df20fa42f553703e6a55a7ce730e6aa565b508cd9e676b2,2783ff7922e7de2fa7cee83e600eb567df40742698601a4653d1786a43930ef0,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,30.8571428571429,30.9,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,chess-puzzles,Chess Puzzles,games,b7a6c914888276f682cfced7fc5353b1889c37ec4dda13d2a8cb03ea676b3901,22856c382e973e3f0d52d6adb53b1ff0a761a1c03ee1bb742ff8cb1d589b51c3,3c040d631f24d8ad59bd5d84fa318c03e503eefbb0a7fbc628ea7f614e16c8bb,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,56.00000000000001,56.0,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-13T13:41:42.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,arc-agi-1,ARC-AGI-1,abstraction,5fb29d09da01e911fc3ed518dff67029d7ad0ec120a9cf13a9aafc57a45d36a7,1529c71477cd7c8ce70e7e47e13bf4136463c9f85782b06449eefdc4f0aa1ea7,d7b031deb2993c493ad55c165aa96a6e265dce5585f86616aed811a10d6834a5,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,95.5,95.5,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-01T00:44:57Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,frontiercode,FrontierCode 1.1 · Main,coding-agent,a0b09988d9298ba09e6455eeeceb3c18b4b9453bd9c7137a813a8547d0ff538d,afbf3304d224c29c81a42f295a05feca766f134787366078729d7ac89b3a1f67,37ba1673ea5b1f77ea7bba1408caee27aacdb2cb0784d7417b3a75b89083fd5c,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,49.27,49.3,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-01T00:44:56Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,ioi,IOI,coding,65463fb2ec2cde32887d52eb6cb97bdec3ea6fff550d8233438d7d554b0249b1,d6d585f7d6367de735f6a904a56451f87f876db1b3c87ca5b30ee7ebb40cd67d,80156255415b1247c4be1849e3b1fb406ca75e82ddf3eea4f1395bcf7488fb9b,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,82.611,82.6,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-09-30T18:53:22Z,fresh,31,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,scicode,SciCode,coding,7546a9ff8d70375c497c4e13da471fb04bbd57945c4c59fcbf457635211c5913,80210d9b0aa0be271172d7f70ed21b181acfaf7898b6afe6d29541b4f07f4091,b1dd6ec6a74cb39581a5891ec66a630e5ef80ec3d83b405f17bcb8c5d6ba8ad5,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,57.63888888888889,57.6,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,ale-bench,ALE-Bench,coding,7c48dacb549def427c32bcd8e68599e4296babf6c78b54d14826b86bc749d14e,4c0dcd04b726dead1bff926b078b3e2349ca2b88693ac48d4679adce0975fdbd,6a34c99c14532b1f26a2789f7735f72f05f0d81749399afa6ba4f8498f756292,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,2462.0,2462.00,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-01T00:44:58Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,vibe-code-bench,Vibe Code Bench,coding-agent,e0cee9745926206e901514fd00c17d0d864fa9f7f8e529a57a41a2339ff81fa1,0a22a48f1711103205aba1f923958db8e12047965496048db5a7f961d00e04ff,88ac255e46c1cc0fd20b4bbb8726af866f425e99124ddab7c2f5ce436ac8e185,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,87.824,87.8,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-09-30T18:53:22Z,fresh,31,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,webdev-arena,Text Arena · Coding,coding,15a6d51a6a8a7553ff3f323b950652be7540b156e5aa19eae0069c67899bee59,a0adb61f9edaa8533c6c139de6c82edac670167984578dc4892ebbd82b797a42,1225567aa529b62bbd6f38b05533dbe341af28360297bd7d34ec05327e5dd885,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1678.189584003976,1678,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,programbench,ProgramBench,coding-agent,361b9834ce7f7f16a15b6d2a14799b9bfca41cbe6aeceb4865a1331cfc94c66c,52b1abe327f6d2168b4dd1256fff69d06c5850dff6c152d07c0acb78b68910f7,2d69e437314107049b8a8f122568434c145e2f95457508769195e6734d362d5b,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,2.0,2.0,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,weirdml-v3,WeirdML v3,coding-agent,9f1d29783754aa23b636cc6a4eb23e0c54d8498432b6e36f9dc2925c0b434b90,d4c01701f19235277d67e004c7df07fa27c0397b2ce44a773be4de8c82ffcc7c,bc6f0a429300b2e2e4db650e3325212c23861b7e6e9d29c5caa45ddc4c53f91c,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,19.731756666666666,19.7,exact,canonical-source,profile_95ca0426927f6f7637753fdd,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,10dca70dcfe4d86a55dc0d6e9d509a4180276282036317a67eda3194b44e9a50,f1d23c1165958724218fe0510ff40551711e58bad011c1686edfd006a56c64fe,fa669073d6a18d088b8ab262478d6f98c7c3bd89d5eb10ef1faadebc265f4eed,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,83.146,83.1,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-09-30T18:53:22Z,fresh,31,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,274eacb219740fbf6d1ce2dd9d886cb5f2d22bcc94707f4e55d12144bf87abd7,7258543ab0afaa91e60abae3f5847e3054b87f561d32904e075ac4494b3e723d,6a9533a9bce842c13ab4f29ccc0b0e5ba4204f8ccb45cdba17396144f587d10e,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,44.444,44.4,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,sre-bench,SRE Bench,terminal-agent,1dfe5f1a04432d2168af5d9a485a0b3d80d9b2092f1c56dc934842a44fda3529,e47715644fd7ee45d7697137c30e74f224dfb86b62819abd915821dd88e4d604,a775b1bc6703fd69152a1c474c83c16a95c6c5d107bed0679b0d963d93b14849,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,30.534,30.5,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,apex-agents,APEX-Agents,tool-use,23b269301e2bac5e92cbe0075e6c8c7158aa48d043a783f9ce8c050c64fcdf30,9e759779f0d14b673dd0252433334d627ed66ed0f55b7d6ac4ab9bddb12b9ac5,5fb09ee23e7bf8a678bf030d728afd17392fb8685d8a87a3c5b0fb2a8c0cf47f,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,54.300000000000004,54.3,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-01T00:44:56Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,arena-agent,LMArena Agent Arena,multi-turn-agent,e2a14bc236153b4c3ecf9cb3eb9cb675c7e40275ad6e256afe817d5132ea5331,1a9aa689d877250955be43cbe0f24ad3a445e911393ab3dbb4c30cff5541ae3c,59d7a985c1839c87874b3ea22d374b32101c9b6bbf15a60b51dd9ff0cfc07fad,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.07955487093669349,0.0796,exact,source-specific,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,vending-bench-2,Vending-Bench 2,multi-turn-agent,1cbd0c84728c7867807fdda58ecc40a4b56b0cce3874295bdce0e7207478c254,6b31cce7f31666869aa6dee2400785a838e3935b3afb01f857e85a7b9e02e410,b9e3ab661cfd42b634c3fa7df2ed91045ec8e64a4189ddc39278315065b437b1,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,14427.848333333333,14427.85,exact,canonical-source,profile_88460deff22c56bc1bda18d5,reported,unknown,,unknown,,false,2026-10-01T00:44:58Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,98111d7c58eb02860f5f30ced617af33c0542bc2cdba9385013a2eb8200696f2,d6ad707dcf8c247fe202d2c6ac68384dc5aca2be2e786214d280874500cc0c9d,96ab81ae4b49713f781b89250902ba8a9d4543057b4d5699500f63c5d986d76d,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1515.88,1516,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-6-sol,GPT 6 Sol,OpenAI,closed,9,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,3b125a955d5600f2d3611b147b635a7f6da106f52f7d9bef883b92f956ae8e3b,5dc90adaf96ecddbf6f44a7f92db62358798509f3102c76eb8093f1cf872f9a5,e2101a01d343a76b89923f742b04b1d735c638a914cb1c533e46805b1b5f549d,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,30.0,30.0,exact,canonical-source,profile_0734f29421da39f59df77231,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
openai/gpt-5-5-pro,GPT 5.5 Pro,OpenAI,closed,10,epoch-eci,Epoch Capabilities Index,capability-index,6f24f643fd41054569ece332736287f2fc304d7fbcd9105c739c1d80d657b0ad,241839233abcff8f4163275390a76ea8731bb827ffaef92a88a30278a7fcbab6,055fcc4b56e82207bf014e80f2b95e69d3ef0484b8c4a1ac93eb9856f114e5ee,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,163.11,163.11,exact,canonical-source,profile_a898b81cd13e7c0a7333367f,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-5-pro,GPT 5.5 Pro,OpenAI,closed,10,gpqa-diamond,GPQA Diamond,science,e5c459e323bc9a77cc9e7c866ee8958e8f29ce1d7a7024dfd2b02eeff32e28cc,94d4706c837da852af9a62cc6f6137513f07c5a7616584ef1ad492c697fa5b85,8a796135e8622378052adc7a4bba7c6aca7c1487fcf5796012bf357caa9a8c02,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,93.92135642135642,93.9,other-supplemental,supplemental-source,profile_52423dacb65ae6d97e85c24b,xhigh,xhigh,true,fixed_effort,,false,2026-04-24T16:01:38.000Z,stale,190,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
openai/gpt-5-5-pro,GPT 5.5 Pro,OpenAI,closed,10,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,901ed971c51a7e914aeb897d5cd9d1f33c949720be14bdaa98506448ae91727a,0b7655b7251b61a6bc23f928d2ac93c3302877e3180abb38cbd0eb7cd165974a,30d980fc451103bcdee4e637b703d5d5481608163460c228f5ae952e8a48d05c,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,100.0,100.0,exact,canonical-source,profile_52423dacb65ae6d97e85c24b,xhigh,xhigh,true,fixed_effort,,false,2026-04-24T23:38:07.000Z,stale,190,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-5-5-pro,GPT 5.5 Pro,OpenAI,closed,10,frontiermath,FrontierMath T1–3,mathematics,21c731c4fdf4aef80bd5b6cb25898c2277e629983ecc1edd38d5718c4ff14c28,8cdf4bce722a77b01cfc475c6adf99c645d55da156a798b1b37340016643da29,4f7bfe326c1ca2ae446294e4eb0445dfa7589ed6b6ed974b7c4a84b746e164e2,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,87.71929824561403,87.7,exact,canonical-source,profile_52423dacb65ae6d97e85c24b,xhigh,xhigh,true,fixed_effort,,false,2026-06-12T21:33:12.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-5-5-pro,GPT 5.5 Pro,OpenAI,closed,10,arc-agi-2,ARC-AGI-2,abstraction,9ec4f8f06ce873587d292566fcd4e9efb850ef7920fe1771b57372ec01242497,dfeca26d988392632d6002ac2c1e0834ca2ff835c804c907fa247095d174a493,5fa4ab1dc988b73dbbbd3e7038300348d2ebb3420f14d96e9a745555386e7758,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,84.16,84,exact,canonical-source,profile_52423dacb65ae6d97e85c24b,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-5-5-pro,GPT 5.5 Pro,OpenAI,closed,10,frontiermath-tier-4,FrontierMath T4,mathematics,eadb2fa4b724fec14345cc7722b0664900b218e48120520ff8a46c3d03222086,66503e0d38c8c1aacfc897e45e2219dfe7f780a0925234f4051fb80f39280d33,3413c81f5619400532d2201f988104f1eb7009ba28c73c1888d3e790ea055219,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,78.04878048780488,78.0,exact,canonical-source,profile_52423dacb65ae6d97e85c24b,xhigh,xhigh,true,fixed_effort,,false,2026-06-12T21:33:12.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-5-5-pro,GPT 5.5 Pro,OpenAI,closed,10,critpt,CritPt,science,5c82b2c0d9a0c98c16452c07ddb851dae9db759aa7dbfe0c033fc66a91381a45,10a8596cbc9f7f66d9b79942c43af9c0eeafffce228b0e49742ee17278a74109,6e9a72ec7b397b00d716faa7db3def3b0a76a46eafd9967d76071b12b785492c,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,30.571428571428598,30.6,exact,canonical-source,profile_52423dacb65ae6d97e85c24b,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-5-5-pro,GPT 5.5 Pro,OpenAI,closed,10,simplebench,SimpleBench,reasoning,f9d8078f687855dc332e5f72dfdf1385389c291a35d420d584c3407f0655dc5d,48be932c458154928ee621e9ef067c9a08105ab221beffa8311bdec4f2e29f25,8253bf31f96ae1c9afe2d99ed7682acbb2c7f46187d7f9ccae405ec8d28f9861,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,76.9,76.9,exact,canonical-source,profile_2ed771acdbb41d4dbb230330,high,high,true,fixed_effort,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
openai/gpt-5-5-pro,GPT 5.5 Pro,OpenAI,closed,10,chess-puzzles,Chess Puzzles,games,5864f6a0362d5b410247534b44a66588ba3676385688fc5f830e2dfbae297b90,57310a8b975e256b0fe6690f32bf83ec8ae3f8b2881ff16adcb63bb35255d489,3051903b77aecca65bd9bcd122d735964e437a40265bf7dc07f8dbc134633314,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,62.0,62.0,exact,canonical-source,profile_52423dacb65ae6d97e85c24b,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T09:56:16.000Z,fresh,17,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-5-5-pro,GPT 5.5 Pro,OpenAI,closed,10,arc-agi-1,ARC-AGI-1,abstraction,8710da348a66b97032b97b958134a7a6e95be688e7146336e73cf3d6a6389284,3af7a9c17d4432ab033f0c7e1d9cb67adbd6a76c9bd66e3b171db2fc956e6554,5c55a0d14a562b4baebbd7babd6ce742dc336a3fbf3083fd945dff3182317256,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,95.0,95.0,exact,canonical-source,profile_52423dacb65ae6d97e85c24b,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,epoch-eci,Epoch Capabilities Index,capability-index,c4a8be40b5cd796772818ae9c1b8e159c6541535f590ed3a0c975d191f9ef9aa,b08757417434b6de1bf317d851ae47759af4a419fbd82db938e06c4a1dccd908,324b4325735f95c565d5a1a034e1ffa643ab72471785ab07045f4222d00cd63c,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,163.02,163.02,exact,canonical-source,profile_d040556a661041609251c38e,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,arena-text,Arena Text,preference,31fe857d42bae7ae4de811e75952bfe51b4aaaa9ed08d35cc3646dac0f342b10,8b01dda50b2f80039d4a344fb16b3297ec4d2614f0fbd2b18fb9a2bcf3e0ee25,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1503.8924288671988,1504,exact,canonical-source,profile_7f9f1833ebf92e99cdabe7b9,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,livebench,LiveBench,general,f7460f69f93a539f7627e27497035c15e9669da7f52b9bdd4c0e0be4e1c9a817,15c69e1219b1415e5afd39663ed36602806329b2ae246d5ee3c3e3dbff9a747e,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,82.97079761904763,82.97,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,348b9fc2f927f141be2832f10ff6a2f4cfd1e311cbee7fa94be8c0b691c04e2a,e41d0e4fa6b05f1fdc571b507b34bcfc035a7e4397bf1413ced61acc76c36886,2b66769f3d0af076fa2b929944b6ca4b078d46e1e57f4168d926863422698241,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,55.468026,55.5,exact,canonical-source,profile_a31432cd65f025d42eaa2c1e,adaptive,adaptive,true,adaptive,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,simpleqa-verified,SimpleQA Verified,factuality,64d30b0ed98e5e453287401ed7d069fe1e8f7ace12119be5d4c13ee8b55d3161,ce4df78e069ee0f8a84bf848bcfc48dcad31fbc3a6503a02baafd3b81722d79e,049ccdbfb7201fed22bb9df66ab7dc0362db3395fe5f74a70bc801bec87c0e59,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,70.7,70.7,exact,canonical-source,profile_d52a4bf6c494a0166f10bbfd,xhigh,xhigh,true,fixed_effort,,false,2026-08-10T23:44:03.000Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,gpqa-diamond,GPQA Diamond,science,d9794a0dd88a7abac5c048bde866169291cec0f0b9ca662d192647cce8141655,82b592c1686593ceca9de9f8b3ececc0bd93f349c8cb32badda1624acff9a1fb,fe3fb05bbde02acd8022f37a2cccf1d48a4f185f7bb091dc994e6f4c743279b8,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,93.182,93.18,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,mmlu-pro,MMLU-Pro,knowledge,606c865b70ffbe1879fd61b033da64b6fcaaa1be5a4685540f5f021b8b697db1,f65eb658fb9a49746906d8d29d943b3def35cc36409007fdc7f3092a5201e4d0,f0eafe717cb93e184f2fbb0c67e69380e42a15d2cbdc22a3c212310a4e25ab74,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,91.502,91.5,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-06-09,fresh,145,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,2b832387d2c5e77130cdcab23c5e2f41e5632d91d131a21e10d53d529dd6615c,6024580f16903303ad6d7f99803eaa177d4b993251ec5ae830a56d80e376ad64,9c3130e76aa52a956a1a814788fd50c9494371f314219961e8c31b3012597ef6,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,99.72222222222223,99.7,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-06-10T20:03:17.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,frontiermath,FrontierMath T1–3,mathematics,ed6aa8345d7580907ca034c9b56ab79516d1d540a2e658d504f0268daa5b7b4c,c929e99fd95a4b559b49015e1b9a0d46f43607bc7eb9885a5dd2c95b800ac7b9,adaccfbd9ea3dcb2d9587a3303b17523448277ed37e5f06e49625bb955229994,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,87.01754385964912,87.0,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-06-09T18:12:35.000Z,fresh,144,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,arc-agi-2,ARC-AGI-2,abstraction,bc9050dc40ffe4292ec5c712bac35291635a433c0ab478faa5f55c8a9739a7d7,86c783962ed5229f01d4a83cad71d2b548a21ab4559d34ffd4bf0ad31692c24a,e024ca81ea8330bd940276ccf8b5472630fa8413936da5e779560a9ef9b5e720,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,89.16666666666667,89,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,frontiermath-tier-4,FrontierMath T4,mathematics,cebae25c4f386b6a24f04d1469d9d45cd8df6c59c339330019e6c5cf3f47868f,da03052d2f15863b97269a475d7a8edc145ca193c153f14db75c3ea9d25662e2,d9161e1a1af891cceac563616939e3f46ecce6fba54d98fa4deab1da7477b514,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,90.2,90.2,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-06-09T18:12:35.000Z,fresh,144,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,mmmu-pro,MMMU-Pro,multimodal,056b0bed7664c884b25eae46cb89245ee5745876849676e52008d206ba32808b,436a443418bb61bf68118de3b87cd9cc97a528275fd1c510f5abcd69d9059add,822305123f4223f467e76361ce1d27e661e12fd04ff00d7dd6a6739bb6637f38,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,89.306,89.3,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-06-09,fresh,145,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,critpt,CritPt,science,1f6c8a92dcfc82c2735af9b2b29abbba8fdb9aa74de1487e15f7ae73f05a8c6a,f4032edccccba4f202a1d96907ff7d2deb1f3decfd3b18f31c6982b8dd1a48a3,4321d6acc6abfdae97c5ed803648924165747bc8b7ffbc9b9eb203d02a56e7b7,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,28.571429,28.6,exact,canonical-source,profile_a31432cd65f025d42eaa2c1e,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,simplebench,SimpleBench,reasoning,5655ed7c02718c2dc1c44b9af2b65b5bd16900e30f32cd2505a6f31877179fd7,5f37c58d88e999fda5598b61e7a96e9a1c356580855b6a707e6ac1c5be78767b,cace3a02d8f4471dfb6c4003910470ea84db930ebd09f70883858ebf1beabdd0,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,81.9,81.9,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,chess-puzzles,Chess Puzzles,games,3f70dd1b632a3bdbe5c86a033a56aa2bc0299b452649b92999054799375e5684,9bdd3724c507c825a5feb0eb8effafbd5fdaab4bd348c5363c109f3c7c684cfd,43db1cd9cc34c69e6119c2d8ac08ce75a756303fd384a46ef2ba5811050bfd3b,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,47.0,47.0,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-10-13T13:34:37.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,arc-agi-1,ARC-AGI-1,abstraction,b28615ee2fd53429cd2f16bd4c3b5eaafb6f6570ff5e5520892404553b77b15d,e78c228ec434f1dda728ed0a68263f517e48adda2dbc29bcc79fbfe48337abdc,b0414f41aaf342e7ff1cd853c181942869f80751290a1146b532d5a39da93666,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,98.5,98.5,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,swe-bench-verified,SWE-bench Verified,coding-agent,641a0f27e970032cb3eae7fa7e6d5cadd37dd9c1ee7a9a73614f3d1105b5134c,c2052f382cd9dfb85645d50ddacde3a03342ff051ec8136ddd047bbf36c509e1,f3f679c38c4297732208f6d5430f7d18296f89b8fccbca3ad7ee937a89ec37fe,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,95.0,95.0,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-06-09,fresh,145,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,swe-rebench-v2,SWE-rebench v2,coding-agent,fba1a8e8e98f75f3a3063fd0bdbd751a2b81a9e024ed34400ad1a5635f53d733,c819e155a7400554ba965b00d4cfefdcaa0893cc6b18d19b1bfa23360fd695ff,cfac8de7b5efb6b9dcee1a89f3f6693f4a7a3571e3f28a7533a9fdc2276816c0,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,64.5045045045045,64.50,contamination-warning,contamination-risk,profile_7f9f1833ebf92e99cdabe7b9,high,high,true,fixed_effort,,false,2026-07-01T00:00:00Z,fresh,123,0,A+,0,A+,7,C,potential-contamination,2,potential-contamination,swe-rebench,SWE-rebench,https://swe-rebench.com/
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,deepswe,DeepSWE,coding-agent,5b1d897a7c8933b8bae3d1c20b81de5d138d82c98396909cae93ca1da4ff362b,be6c80ea3b852479be8ca41f35808eaea58ee510f26dedc16d021569f3c6fdea,797587c900d28819d65b1a843e000619b9085a9e51651fc1d82f041d9c7727b9,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,69.72477064220183,69.7,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,gso,GSO,coding-agent,f7f58d9bc83a727e891f76aa0e73d98df759d25b1cc23519f79f80e9f59c9c85,dd7170d127d603f389104988b09d45a96a548fd5a6ae83477ef1457d90ee52a4,314f8356da5124ab422275fb0fbc03290e1c215845de19928ebe8cd590f07d2f,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,78.43,78.4,exact,canonical-source,profile_e25453435ebb1777deabb6e8,reported,unknown,,unknown,,false,2026-09-27,fresh,35,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,frontiercode,FrontierCode 1.1 · Main,coding-agent,3a8de3d7aaf29402002789ef80fd73da1e0373192ed3e4f018aab6804295f90b,3d3f4b9471aad9b4afaca3b5c875956f2b9a597cb4be02aa464a5c5709d1cb16,6649cc8bb042759199d52c18d26911120b5def9ce97292c82302590f034245eb,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,53.480000000000004,53.5,exact,canonical-source,profile_d52a4bf6c494a0166f10bbfd,xhigh,xhigh,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,livecodebench,LiveCodeBench,coding,612136ad236abdc9a8f82670f04aa2d3ad352ab7374b05f76cd6cde5f9788762,5139dede1544028a465a38ed57ea2c5ea44f468377a1c0cfd484eb13e62f041b,187a360b796425992516b601a57c2f1bff0c12ce6153c655c3d651ac69408342,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,89.778,89.8,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-06-09,fresh,145,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,scicode,SciCode,coding,f96f2c061e9f5dea51e38cd08db8bb7ef0d6395200875a030f9bdcb866ba06b2,0bbadcffdcaaa03a3d9cb5a63ec3fb517c1a10e20cf704109bc603ab67006cf1,e0b2fcc0787cb427754dd973e5b44c18fa280fdf3ab70a7c7fd0a385185b901b,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,60.9953703703704,61.0,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,ale-bench,ALE-Bench,coding,d6d6e978ae1711d45ee87f229348faada3d49bc0af523f683ffe600fdce3332d,370c9da7d673bef9427e32e4baac5bd978fffdff380d06232035e8659a4055a8,dc8244642f7f31390dd20058e4f48b24ed9d0984f3ccb1e5aa415514fec7f53c,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,2041.3,2041.30,exact,canonical-source,profile_7f9f1833ebf92e99cdabe7b9,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,vibe-code-bench,Vibe Code Bench,coding-agent,0d93757af2f6546896262f85eea3b361881535aeaa997492dfc7baaf35d6e865,cf2cb69917a609e7faee2c6c9fc26da410e05353138cc06e97b033f99dc236e8,5238afb805eb35940ad63a588a9b8f6154ee5957c0b214bf4d0e77e7c79a90f6,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,90.352,90.4,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-06-09,fresh,145,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,webdev-arena,Text Arena · Coding,coding,7279f34d75e675c9dcc09283d88cca598f87777f2d86b9522d8347bf62d74f27,3f812c4f2d20c1caeee18f20c16d86b87656f986fb21e05a493afe48e9279770,9f5506b2ee113a17f88afb721cca43bed0ff31c761a8eb2eba8b8791d5637e8a,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1624.082589402146,1624,exact,canonical-source,profile_7f9f1833ebf92e99cdabe7b9,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,mirrorcode,MirrorCode,coding-agent,086adf5fe9ebb9142cf1be3cec72569eb5049a8fc66d93a97d4ef9d1f1b1c54d,335246bd498a881e2e69787d33857e5eba3d1c8be0d99e2dee5600f6ee4a7c87,a41ce8d08c70f058547b71d087fd53b98aaffd997c3316745d4a2fe42e4ddc5b,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,63.88888888888889,63.9,exact,canonical-source,profile_7f9f1833ebf92e99cdabe7b9,high,high,true,fixed_effort,,false,2026-08-10T16:12:44.227Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,programbench,ProgramBench,coding-agent,3690fa58346cc3de8c0f6c35a7ccb4339a1677cb549328cdff0719c6ae4c26ea,3eadaa8e9dd6c786a225d5f1ee2c00e3ebaedb0755d5b98d6800447c50834d39,a80a410b47671a3bca6770bb811c696be5737a6347cf52551bbf98e70f127b47,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,2.0,2.0,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,weirdml-v3,WeirdML v3,coding-agent,9f07d20b4dba0ad213cceaa382a8d6b31717c372ea56d765a4eccb81cbe8abf9,29e0616d95e0fc9630037351b294fdb14c9b81928fa7c6645d5de4d91f4feb16,14eb5a8f690d40554db17277cc54c53e11418051175176f6b3e0c798f451b527,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,15.586933333333333,15.6,exact,canonical-source,profile_d52a4bf6c494a0166f10bbfd,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,6cca0e3a967f665dc49a23bf8c2f7c14bc6312fadeeca74747b41494a2ac881c,b4fe0c6b38248412eca8b3c078ab32b7a79a9ac87def38c2cdb0ea0f8bd9860e,de0e7b75cffdd9188f364c94fe0801b76d661dfb9410d3f8cc226b4838f2fd40,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,80.524,80.5,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-06-09,fresh,145,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,1e12d65fafc8b599913c7e7202292f8127a88e7a65cf0a670c890379ac880c7b,9c3b0de1312235a5015ae8aaf86e6bc491c2af84651722345b548841561bc4de,08001d342ba8886a610a6f719c17c934469d9da35ed21be61b993f67a76d1316,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,41.414,41.4,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,357fe51c2a0d5e66b8c6c76e7191fd836c0cc5fb0bfde8e1f6c28189150abf3b,456c888cbf658073978a21446424a83f647cf48c8f079ba2163f141b9d29a973,e4adb326a2e894634f4ff7a2de41020240efa24c3a32d5fc523ba0d4d01f753a,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,38.14433,38.1,exact,canonical-source,profile_a31432cd65f025d42eaa2c1e,adaptive,adaptive,true,adaptive,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,mcp-atlas,MCP-Atlas · Scale,tool-use,fc211217c513e2d76d728421d636bdb4ccecec7f38998ac7ac7902684161eafe,c1fe4bf26c8c1a12f603c94adf36fed68b2ed305fe556fc75a0bd80880307318,855a8158b20ecaa1c6431b2eec7b29dfb95bd47cd81dd5a6433a5aafacd05b1e,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,83.3,83.3,exact,canonical-source,profile_185ead53e002c3c1ba45c533,reported,unknown,,unknown,,false,2026-06-09T17:51:31Z,fresh,144,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,apex-agents,APEX-Agents,tool-use,d00a31214a47a7786f92de3f34ef1ce0fec57b62df7efd09807b8693fea481e4,8f1be2860bfca6dada768506ebb7539706ed0eaa5ac18c7d941f455a522ad965,bc9d110ba473cb81753e768dd9fd0b3af54ad95a51dff0f750af7c5449f42f6e,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,63.6,63.6,exact,canonical-source,profile_482e02f0349d00ca23df8127,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,arena-agent,LMArena Agent Arena,multi-turn-agent,d10f60f4ca8abc579f2fa28eb3130f52c1453f6f4dc4630745995db3c8d1046e,01cc162403b81fd98c1cd5f76259694a40405a2a68e68d586e2d7b840e261472,b4c953ead6eb790e90823c14ebc6e0ffa52b04ab7304a88b7cdfd2431c31dc93,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.08037071257211915,0.0804,exact,source-specific,profile_7f9f1833ebf92e99cdabe7b9,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,vending-bench-2,Vending-Bench 2,multi-turn-agent,4fe7edf7cfad33cf9f2f752e83650ac94c82d550e904bde32a91bad5e7939300,e759c9055f5b5f1be891748b171881eaf01ec0f480c551f95cb854bf1845ea25,e24bc75ebde244e669d9ef0cced07b951a49d50bd35095b9e3d95253c8198b00,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,4966.639999999998,4966.64,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,77f500e3bc81018c55f2960bad09d23fd915c898cd652d579f4d0a69a593585e,3ee5f8f0e3bb5cb81bd066d7d9cb5b2e2b91cf8acf999b0fa8570851ee579d64,2a57587dd2216d99248636487bb8135182d6c3ec78357b399ceb07e09ac23b03,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1618.39,1618,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,terminal-bench-3,Terminal-Bench 3.0,terminal-agent,4a3bed6427323f5a1cf334d3ded74c3caaaa3c1d033cac8072648dbcd43c4927,5318b343d3979e9d6380768d401b794c0b627f039360f56c1f55963367e2e965,3d11a4e7e9474b518de9518ea5ac6bf9a0167a6e45624d8bb01ff8cf092ddc04,release_ceb36f25882d8cea2c9d084a,3.0 · b936386b-4afd-4ad7-aa7f-6b52bcab9853,official-74-task-leaderboard,accuracy,percent,true,34.05,34.05,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-08-12T21:39:34.661007+00:00,fresh,80,0,A+,0,A+,0,A+,official-leaderboard,4,official-leaderboard,terminal-bench-3,Terminal-Bench 3.0,https://ofhuhcpkvzjlejydnvyd.supabase.co/functions/v1/leaderboard-read
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,41c3e092017d78d6eee712678054e0367de5f4d54131a86239c73b0268d6ffa1,0987b684e171e36c46c54380e8d966753541fa1a7ab768819c6cc118d044d22f,be9fad9622f1f12c7ab66b7b1d9d0aa23e2c20fddf0fcc4418b0a7a409679190,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,15.714,15.7,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,60b4a94b1b4ac2871f182981698f44b98c6dde5860c32db82aed5f929d8bed4e,f5547c68c93087bd8a320bedf59e0a9fa23a0946cedd05d0a0cb0e0193ea4596,bc63906b45498a36c7f69245c165cd52ba114e01c9ead5912d9063301c0508b4,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,39.69072164948454,39.7,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-07-23,fresh,101,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
anthropic/claude-fable-5,Claude Fable 5,Anthropic,closed,11,weirdml,WeirdML,coding,8f82af622d0a55008b0de0265ef063550e105d22b906fa0f947ac98b1609f611,4ef9b6993286921c6a9702f8ee8e8323cbec3889360455b8c0bbbed825fe8b87,b9e36ef7392f80cf55925733544effa6dfb854521664ff9ded2b243e4b70e2cb,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,91.94,91.9,exact,canonical-source,profile_c6c9b7c96eb814bc15191e40,max,max,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,epoch-eci,Epoch Capabilities Index,capability-index,bdfdd232d2d12f3c94d3d53b4100a04dfe4a078c5142c2650ec9035c1bdd110d,151f3b1e4662834b8fdeededea8e441f278b85035888e081c7d8f5cce8bed431,c8e5556998763c108258ee04b24a079cd62a63ab7da86e135aa4b94c37d4944f,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,162.55,162.55,exact,canonical-source,profile_7ff85e71c65ccc9bee850b2e,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,arena-text,Arena Text,preference,abf3b554860cc56b380cda4b63e65bef9f83485ecf2f2a16abdc86f7f76f6564,6df5e16d512312537c83c224f1ddd989a28f3b646002c5946cce7696a1a93dc9,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1484.910477373644,1485,exact,canonical-source,profile_1b56e52cdc1b51fa09ca0421,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,livebench,LiveBench,general,70c9bba381113e56bc436fe1ed16c60e3f71f727f43e27e9e4cbc823441554b4,e090fa41cdb113a42a2b6e22193d9715364d9b8c52ea99160d82a0bf7a57764a,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,81.05354761904762,81.05,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,5eb71b21285c79813d310dce0fa1951955a9109b5c25e61725f9258ef5dfca7b,e5ccf8a301b169e92a710894f9860495b02cfafe02e6cfc3e8dab7a1879b3afd,db96fb25810622597f236f437c027d00a5acfa3d1c917e4ea2af57ad594b4bc6,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,49.490269,49.5,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,simpleqa-verified,SimpleQA Verified,factuality,5160c9f92fc610292f547001a575a81613399d21c58ea89aea8ab0a5e8a3e358,8929324d75dec5002f8ae9600f99349c0f7160ad3430a65c3930c6ab69d6b151,e8c442b4dfa0deb1c59f43f7f6c56b371e45a3134beb26cfbd35e2d47f70d096,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,69.69999999999999,69.7,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-08-10T23:44:30.000Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,gpqa-diamond,GPQA Diamond,science,0335b5d5c743507b11f4e7cf953b33cbfe5e35c4d6ba4b0fedcdcd07c3db1883,fc6f12c41fcc9a2064098aa477461175e3f10fdba0639ee7c1e51bedba0db320,4bddc35537230abb9031166776429e9322e24a865a25aff9070109b412c02d6a,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,95.202,95.20,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,mmlu-pro,MMLU-Pro,knowledge,b5d67e4b208802174570d52b4ba2f54f0203a93b685fbd8e78e6e8f7bbce1341,7e8e470af07d50d63b5d33325e590a700feac08681e32f560cdf3e8f970fc36b,13bf06df9f9d63ea18257440215dffe6d85a3acda2b0a37fd602cf7961aea7ca,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,89.1,89.1,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,293a503b0cff38362d4d82edd74207dc3f41e9521a466439652f4780e2fffd73,7c880960c0d8c6d1ea6e29f15d9d328bffe44fb922ef8e98a60c08b8e3c3cb6f,af444f0831bcc1693c151ed626c6f9124e9ea5e83544fd5e9442f46ec348b72b,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,100.0,100.0,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-07-09T02:07:13.000Z,fresh,115,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,frontiermath,FrontierMath T1–3,mathematics,8f3979d351455e10edf3b7b6f9f5a66a9cd4191d8e5d18f09141c1a8cb5941c5,4b16fece70b6ec62e4607f13e886804db90154a11f68c3c494784ae7f6ed8659,fa566697fc7062a3f4e2935fd2d9854f3b2afcd922216eebe37ea31e3d64830b,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,89.12280701754386,89.1,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-07-09T02:45:09.000Z,fresh,115,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,arc-agi-2,ARC-AGI-2,abstraction,ebcc2c9064c24ba1302febbff1b5702543b708c5babaa809986f7f93a38fc549,e279f93c86a0855ff81c42d7e263599ecb9a760dd6b24259304aec009d9af800,377ce036d8d437ea9dce0284130749a9714e2195ff5b9f063a11d87a867c96e4,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,92.5,92,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-02T13:33:02Z,fresh,59,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,frontiermath-tier-4,FrontierMath T4,mathematics,f6967b428c569b07e8713bd630a1b26032144b8960485883f03b8a584f60718e,8f7dd20c587144e337400d95702c586418723f546a0f63c22f85e12d2459ecbd,aef2c1d1a352ae61e4d0bf1c566c81c6a875a6e899fa72e7197aef2bf59b2c49,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,80.48780487804879,80.5,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-07-09T20:34:41.000Z,fresh,114,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,mmmu-pro,MMMU-Pro,multimodal,93640fb7b5b2937f1f3a2eb7cd96e4d1d008338ae1599242bad7dd5d157eb95d,6a9cb22b487d24032aac8bd62d5a91e4bf169ed94673483a8c0615f81149b6a5,722e2d51968049d4abf1ab172e834323570a2474c4369f894575539aa1bc75bd,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,88.844,88.8,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,critpt,CritPt,science,1db47950e0dcfe970e7d456bddc9b1553c715c020f547e2eb249b10df160caa4,6ac99c4989f0e451f3e5e65636f5ce02522c45ae8ada1caee7022f6b78d51f3d,e048592ccf101949136e65dc219058a7e2e1b81302bf05bf8f6d193d0ec9d1cb,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,32.2857142857143,32.3,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,simplebench,SimpleBench,reasoning,eed4db70935bec0447d7ae8a6ccbbb99addcc921831a212e32759fcb370557e1,b178f827c4578a7a97376deaa830ebdd8b12f9a2d6caee817294a7325c610364,f757eca0faeb48281cd4178307e2d0dce127d281f981d358abfdb43cddd15570,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,64.8,64.8,exact,canonical-source,profile_1b56e52cdc1b51fa09ca0421,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,chess-puzzles,Chess Puzzles,games,770312960fcfffd5ae7cb335b2ac9481ceb5762646fa2ea755251342be82d41e,39da4e0f90b55cd168c38cb5cb59e0be32e7c2b9927a29f2d7f9ba76999dc463,ba490b4237455700d3ca94f7f00da4bcc0b9dba1ac87266e2dd21a86b7f1fbc4,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,59.0,59.0,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-10-13T13:50:12.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,arc-agi-1,ARC-AGI-1,abstraction,5b08caf0ed1fe8e60fdf47bc11746b8b8dd435949c01874e2dbb75797411206d,d64d66ed5d9ebc3ba9f1350ddd6e978e5bf3b57021b013ee98d6188756006ac8,ec9e359c509b8ca4781d0a7f598536b151c1e22afcc18af93c0e4413ebff89f2,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,96.5,96.5,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,swe-bench-verified,SWE-bench Verified,coding-agent,a8e05ed03fa73ea1b54b5f42ddc930dce9d5f8530c712fc027c0e57b30b4d1ad,d93df57625e3f37bdc16654afbc48acd994b870415fbc11fd09e7c8decc42222,ad9b439d1c65794d955cdc4ff3e4a7bb1e79fbef405d3fb9234f06290360a001,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,96.2,96.2,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-07-09,fresh,115,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,swe-rebench-v2,SWE-rebench v2,coding-agent,50975ff78dcb4886883ae6a8fbd9a79df23e9f9a9483f8c461f8837e939da1ad,041f5582280837b3009bb70c48ba8be0b3667954e59efb13e41b8aa67b072567,d16ae687df2d5dd76d6c38924008b1ef9c7b3d1d6fc457424d2d15395e17086a,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,62.34234234234235,62.34,contamination-warning,contamination-risk,profile_05094233109adfe43508046c,medium,medium,true,fixed_effort,,true,2026-07-01T00:00:00Z,fresh,123,0,A+,0,A+,7,C,potential-contamination,2,potential-contamination,swe-rebench,SWE-rebench,https://swe-rebench.com/
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,swe-bench-live,SWE-bench-Live · Lite,coding-agent,521d8ed5899146a6b0b6a20cef7ab059f056dce76e5f0b81c9a2ded4dbfffc67,f1cfea9c0194b41b9ee6a5c929b19f552f97720866aa1e526c41a51f5e20d03f,bd7ad59d365892c2706fedf0a6bdbe7910b502a8d50f3bffcbda83cdd20c5567,release_9d2de9d5be7e9822af19f7b3,SWE-bench-Live Lite,lite-300-python,resolved_rate,percent,true,70.33333333333333,70.3,exact,canonical-source,profile_4aea036993309752d798d393,reported,unknown,,unknown,,false,2026-09-26,fresh,36,1,A,1,A,0,A+,official-submission,3,official-submission,swe-bench-live,SWE-bench-Live · Lite,https://swe-bench-live.github.io/reports-0605.jsonl
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,deepswe,DeepSWE,coding-agent,2ddc513246078876bbb8f0dd3ecbc4fb0805cc1d5a78b1773874f253b5429e37,367118f965c334b7db4546c11f83474c41c3f1666cff04c2e78b63a752e82c30,a712907929a4595a3adce2f6ebe5586fc1817c713049c58fbf69d2094cee7ffa,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,72.66666666666667,72.7,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,f010fec5892fcb48f8b5cf68d81dbbde1bb6123225464a148ac9c4c026b4b9f4,cbcbaa72c8344cb90a10d7efafeb3ff60d633eda8eb329caaf52f0e8fc44e379,d72f88aa94d4c9ed32ddb455339b95419bbc9839cfa624499876c2855b896175,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,95.5,95.5,exact,canonical-source,profile_1b56e52cdc1b51fa09ca0421,xhigh,xhigh,true,fixed_effort,,false,2026-10-07T01:44:58Z,fresh,25,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,cursorbench,CursorBench,coding-agent,4ba1e9f7ab6380d36d1bf0d4f6d34d497cf3e70d2990ea1f7c7076bd9837287f,712b7fc0d98ef601bef0a9b1ba4bf31a5d896214f3c0a1705d4621d0676fedab,5084ccfa853e6ab353f6a9b29cba903f5d14ec61d985f8f20cf9045c0f774061,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,41.7,41.7,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,gso,GSO,coding-agent,caf4483a41f899171f50862a43a0482191af464657b6d90c4903e63a9af55824,7d11cc30ef56755f51347809f41eac3bcc600123e2b05f124cc1be4ea87c2671,d56a9e3960eeac9835fe6bf5d46fec8eee07c6f46714ca5e4375d7cee7786f97,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,76.47,76.5,exact,canonical-source,profile_965070fd41fb1df7f04b6bfa,reported,unknown,,unknown,,false,2026-09-27,fresh,35,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,frontiercode,FrontierCode 1.1 · Main,coding-agent,a5856d977fc9add892e75efc436afd14c043a09f7ffccf288801cbf848af7fe0,5ca9c0aec3b4c578c2e60900b63e75c99ac2741bbd4eebaa5634c0117c2dc808,0b4a3eaa2469018bd79fde26e6816f088ff1492ea6488ebd93af3954346ad7f7,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,47.49,47.5,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,ioi,IOI,coding,404fb8ceb82c5c05b8dd6c36049d2b8076209296e997ad336021fdfa18f26748,e1084ada55d3042efaa5685252129bd5af3c7933fe6406632ecfdb2f34a962ff,1c161be4cc2f14788a6c93ae6b0d835823fb90da0dd87a01492df42dfccfd9ef,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,91.167,91.2,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-14T23:34:42Z,fresh,47,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,livecodebench,LiveCodeBench,coding,d5e1156986042d89257c34a7538e4d4a0769378ac2096bb09a5b5737f6cd7e8a,26f8a50d9bbd7812d523958651682ffc8444a211769bcf6a8d338fdfe6e5e490,4f12fabe8d2250a5013fa45059236c777dbd88eb8f16da1a50e9d9b4da791afe,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,82.604,82.6,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,scicode,SciCode,coding,caf193f9fba3f365fb2efb8f7087af8029aa8a5dc795b1bfc95e455979017edf,0d887240b1b9fe322ecbb2a0bbfab684e786e235837f8f74342f10ec33bbb344,e778c20895945a18ad949ce2bca1dab91cdf5e768fb516689d40cf333fd5f1ea,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,57.0601851851852,57.1,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,ale-bench,ALE-Bench,coding,3c4057fe05624862ab6a0312e53e399523a133cfbfb6828f93035412440c3e61,52e36433e72d7cd101b5cddf1e6ac66cd0ed82dff4353ab3fe888d234b06d699,66845a98dccfd2890bdd6dca1ad15dcc6200b9ed34f3c49cc1f835106236352f,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,2176.88,2176.88,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,vibe-code-bench,Vibe Code Bench,coding-agent,02b9ced8a7f71096c73e56a671f9f037c089aff56570f19d15c073d3299a4882,30db70b8a7fb3e401c7fd6744cc036ac4bf1d2111e73bc736c35b5bd6fef1b10,bd1a4d2e22eda48391322e1793c5cc84df8bf558eb0788197fa2a1bc88d20982,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,80.495,80.5,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,webdev-arena,Text Arena · Coding,coding,8714bf16870123fca59129638678a68e0683dd8670b03d4cab60c485a2377e99,2c042c193e76e57a716f296234e0195cc90488c8c8bb8f8c8a5433aaff8d8d47,20b320810c74e63e9300a28d0409ea168553d88796b2a11703644b28f1d60729,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1615.7849726003753,1616,exact,canonical-source,profile_1b56e52cdc1b51fa09ca0421,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,mirrorcode,MirrorCode,coding-agent,c76dac7acead3b3eb6fc7994ac6b9e74238f75c222a3310e1bb52bfc8e2699c9,d2d74ffcd4d9f1e7db15d333c77929282bdf7660dc2458d140c8b1a98ac679c9,a58de79cf4f660b386289bc562f0a44ea06f3aa39e01544289220a4104beaad7,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,20.0,20.0,exact,canonical-source,profile_79938598fdbb1914df59fc3c,high,high,true,fixed_effort,,false,2026-08-10T16:12:44.772Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,programbench,ProgramBench,coding-agent,1b87a1fe804b866fa99b4f67b1d92194e60d4ecc2fff9459324c8f61bb5b238e,ce9fc75b27b079e84cbc21729c572b0cd2002275a3a7487d29a3cd6b07bf2721,47773af8a9bc25d41b7fe90a2b6e479d79ccb61a16b918edf3d3361677e4d0b8,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,1.5,1.5,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,weirdml-v3,WeirdML v3,coding-agent,67ca79ded5feb37cf64da2da2adcd4fc9248116aeaa9a2d6e28ead6495226af1,d306c8dcca6d7d36388204d2c92bff990a85bf37d634430545a26701ac8b0109,0bba1ac555db26ce37fab8ff63bae81dfdaf320ef5b0963edce65a06009ae77c,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,14.992224545454548,15.0,exact,canonical-source,profile_1b56e52cdc1b51fa09ca0421,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,f21aca5cb9fd73a2fa740337011f4e9a03afabde1205ec813c8df80db152058a,f3c73ec9ff722a9d7b0c13637d63f9296494f87250eab086b5f86bcf2de10f10,5a4ba5f568b886433683bc8464125fba5692972e726d46b54e21ec8426cbf204,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,85.768,85.8,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-07-09,fresh,115,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,39a6de31eb642b8200d60da9e5637aea2d65724d8bccdf706fe0d98c14ab02cd,a8a33df9fa62680d0bcc7fdbc1897ae055292c8994153405ef961126cbceae56,9257b66f4993e42cea8cfe756a50cf1c6a910cd98250ba6a5b9b21b971ad30c6,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,37.879,37.9,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,sre-bench,SRE Bench,terminal-agent,ce75cb0ac0fcc9516e35cb78ce1f53b3abfa826e53d20c56bcb4d74f0389623d,3c4b061cfe29bc7373492fe3a20e464456868625b1d9f962c7d694a455eaef36,c324258a209eb38ee8b49c1c716e13c95469b96f5c905aba4e30d5705eb032f5,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,30.534,30.5,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,eec93c68dd5ce6cc8d11075e63c811c219120a3159dc07d0c81573ae900b8d5a,583b249a195123635fe98bdd0babd8b221210898e360287e27a0aa21bc0ebd02,23ade084fc38f7474bf05c58f2c4fef336a2917680c322d09980491a27aefbde,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,44.329897,44.3,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,mcp-atlas,MCP-Atlas · Scale,tool-use,60682b4bba16bdff3e962f05e72eea8820743fb27e690d63a87a33bf22e03680,adf722419f931227bec6c1d6181a773b4a99d8587a56282fe6c01e93f9256388,409f1ceed72297f4b148b18ce628ff5efa64a5dc1e0860def74a5eefb7530e1d,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,81.8,81.8,exact,canonical-source,profile_7bcb5fa02e3e66eaa72b4761,reported,unknown,,unknown,,false,2026-07-15T19:04:43Z,fresh,108,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,apex-agents,APEX-Agents,tool-use,1087a1b2d0d5ce1414aad1e5a0de8919844e6866cf4de25aa4451944b5221118,90b3ad1175ddeebfa02c5781b1a7d16eaf6ea70cd6edea0e1746165af040faab,a551a6bc117a9e22f47051d86001b74a59a0ad7ea702f3f3b4fb268d87983bed,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,51.4,51.4,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,arena-agent,LMArena Agent Arena,multi-turn-agent,d18135883c0bbafe1b5211b7eed3debe5a2e6fd693b8db434fa00edcfae7ba04,8e2c84ea64a960fd2e6346ae30be77abf8a611c69f223c45ba211d3ec863acdf,69a242e74339a17be1e03ea77a55e848e882042879a26593cda2d17331b38012,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.04790625499047256,0.0479,exact,source-specific,profile_1b56e52cdc1b51fa09ca0421,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,vending-bench-2,Vending-Bench 2,multi-turn-agent,eccfb55849daa3e676c6565438f04ccbe08b771717962007bb8489fba5101779,b41a28c041128ac5724f151ae9ec4b48ff0d68067a4c570241323bbb2f2b5e53,8a79dbbf928d612db6c65ea087a14a0a1840500e78ef2f83840fa5bd8ad54c38,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,9619.367999999999,9619.37,exact,canonical-source,profile_a6e3048e0b5def1df0f81987,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,04c6d8f13ee59d9b42920a8106c6c88453746e7262c8abdc6dca444556d9c672,ad98070c82a1804bc70c8062cfbe6698baf149fea00ab62b7d961209ddd1efff,c92653d70a8c89a293b9ce89481917d7769209238282863c8c436dcba752d99d,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1617.83,1618,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,terminal-bench-3,Terminal-Bench 3.0,terminal-agent,bffc17539f3443c97cf8012594e6c22586fbc437fd27467abbb4ee17995b403a,d969f0974e04e7354172c44eb11514df49b6c9e5d131fa248e147ad8d3e0a59e,b9bf808dec50df0e64e3025aba0c983c695bea80fdf2f962997e816e5bc5d50a,release_ceb36f25882d8cea2c9d084a,3.0 · b936386b-4afd-4ad7-aa7f-6b52bcab9853,official-74-task-leaderboard,accuracy,percent,true,34.59,34.59,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-08-12T21:39:34.661007+00:00,fresh,80,0,A+,0,A+,0,A+,official-leaderboard,4,official-leaderboard,terminal-bench-3,Terminal-Bench 3.0,https://ofhuhcpkvzjlejydnvyd.supabase.co/functions/v1/leaderboard-read
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,fc3eccc5ba912c67ad25688d5b8afb4b2b698ccbc71080f0b9c34c241179dfff,da52242bbb7f5d1452f280085a6be060d99ec52fe6ba9b4cf1670a375b6c38d7,7aacd7ab587b68ac9c2610692316aae6ee7fa6dcfbe88915b7d978ab95e07298,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,20.0,20.0,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,osworld-2,OSWorld 2.0,computer-use-agent,7a0c33214378f1b3406a22e6c39e4b605f13860f49ffbb9b7a795656b6ff2d5b,ad6b1577820cfaf53e0024915ac51ac8491f0050cb95895ea77f7c3dc2d7b619,596dfeddf4f879c97fcbf0ae4f906947b4af3f10abbe2ec8c0e260ef3c506b87,release_15c5525fe39f33bc181dfb69,OSWorld 2.0 current leaderboard,binary-accuracy-500-step-budget,binary_accuracy,percent,true,27.340000000000003,27.3,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-02T21:27:40Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-osworld-2,Epoch AI mirror · OSWorld 2.0,https://epoch.ai/benchmarks/osworld-2
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,9e2f3c32a9cd48e71962f9cec95a2e54d413996e7b8105d78bc013c911ff9256,685807b51284fb64664d42ea68f5c63a653866e6b50c930da8fe9c8c993b9145,330a2d901d6208787dd59902438b274fd530133a84c8829a19757d7db2052656,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,46.90721649484536,46.9,exact,canonical-source,profile_1b56e52cdc1b51fa09ca0421,xhigh,xhigh,true,fixed_effort,,false,2026-07-22,fresh,102,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
openai/gpt-5-6-sol,GPT 5.6 Sol,OpenAI,closed,12,weirdml,WeirdML,coding,d9ab75610082361944b66eca3888724821fb1309d9df299b199d57ef8d5fcfb7,2b091f35e66c7ccf56a88d5e97b9733dc402ae04f7124c09e685eae4048e7c4c,d10ad421cd3940208257e8f5ba0b686f73dda194844159ca82159e453a2ce7cd,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,89.43,89.4,exact,canonical-source,profile_ae20fb1272fc841ce5c8675a,max,max,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
meta/muse-wonder,Muse Wonder,Meta,closed,13,epoch-eci,Epoch Capabilities Index,capability-index,20c73163c9766394027ce8630241cfbc29b56978ca402f788ea7be0993ec4bce,fa1051c72ff536aa94bfffc893767847dd339e9aa64329c791439ebb93a63a4e,2ae33513ef33f957dcb2b2d5103bc88fe29616a8580b08769560afe3f31aeae7,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,161.71,161.71,exact,canonical-source,profile_2e85b29d500b3921d8c3d7b7,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
meta/muse-wonder,Muse Wonder,Meta,closed,13,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,3fe9c58be9b99feb47311e9a0b4422d2480e758005264ae63c602dda89b6aae7,76f6f8032d1acc2c5ce5fa6c60513c1917f0b74b720a60a806bf481e0545965f,cde2a5b12bc4fe6ac3a7da79dff414421b06faa09cc40077b25a573723426899,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,43.324823,43.3,exact,canonical-source,profile_4359bef657d5e96da8d9989d,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-wonder,Muse Wonder,Meta,closed,13,arena-text,Arena Text,preference,84b77e348079884b0d9d2bb41cd5e87969f45bf6d65b2997ec5d7bc249214aff,ea9b2e3a813aa02e747985212fd22995404d2638df039630a6bee67fe4cdae4b,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1507.0976650433042,1507,exact,canonical-source,profile_4359bef657d5e96da8d9989d,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
meta/muse-wonder,Muse Wonder,Meta,closed,13,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,20db8b7999029649a0a493f0d2b530e70eb3f05b8d0773a887b063218bb495ec,123c9d5cb8ed343fea5dc4c49d76ed1a636c77ad82d006abfe7174837435c98c,9e8a3f4b2193e976309164e99245a733e7b2b43e3e2ba016d3c395062533885c,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,50.092678,50.1,exact,canonical-source,profile_4359bef657d5e96da8d9989d,max,max,true,fixed_effort,,false,2026-10-12T21:04:58Z,fresh,19,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-wonder,Muse Wonder,Meta,closed,13,critpt,CritPt,science,1ac376d12b10387ceecef5a39533c98a56afdd3902a752d51168994a346375bb,e98c591e381f33f6964bd43fc5c7cb6e07a569ef6bed49fc551ec97e156c2533,e4c1076c584d0fe77418939ca3b4cab0f77c89b08ea6ef8a8f78abaeb3ecb7f6,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,28.285714,28.3,exact,canonical-source,profile_4359bef657d5e96da8d9989d,max,max,true,fixed_effort,,false,2026-10-12T21:04:58Z,fresh,19,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-wonder,Muse Wonder,Meta,closed,13,ioi,IOI,coding,403ea6e3ca3c1950c15ff58a926bc2c50251b7155fcca881ef1a8880727c5104,59b6986afdfa58e6dc253361ef58d3507d713772e693ced0b9737fa41a7ea05c,f0be9a5926fe091eabafa70ac121a62734b7f98bcbf7fb74d2c24a155cd4f4c0,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,66.944,66.9,exact,canonical-source,profile_4359bef657d5e96da8d9989d,max,max,true,fixed_effort,,false,2026-10-12T20:29:34Z,fresh,19,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
meta/muse-wonder,Muse Wonder,Meta,closed,13,scicode,SciCode,coding,93fc9939503b4a38d876355ec033ffb5eb52b44274a88003d00c16c93b5e4670,4637d86e5e263158e8e020c9466428c9caee928f7d180bb2cd53ca7139d48e20,2190a53d9fff6afcb1e3a6bd83b83ece697aac8a802ea9bf5ce396832118a69e,release_de40c66bd40a7b5483b40f9a,SciCode v1.0.1 (Artificial Analysis),scientific-research-coding-subproblems,subproblem_accuracy,percent,true,60.300926,60.3,other-supplemental,supplemental-source,profile_4359bef657d5e96da8d9989d,max,max,true,fixed_effort,,false,2026-10-12T21:04:58Z,fresh,19,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-wonder,Muse Wonder,Meta,closed,13,vibe-code-bench,Vibe Code Bench,coding-agent,cf4cc0c8bad2baa3633085900b4cb637a84e776f97b4ed35d284ecc8bbabecf6,07dc938afffdbbac12d6b53d0d127f4bde991e029e8450082f7b107fa995d980,9fa0c9883e0916f8db7267500b9cd2c728f736c48b515ad8a316d25516ab56ea,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,83.089,83.1,exact,canonical-source,profile_4359bef657d5e96da8d9989d,max,max,true,fixed_effort,,false,2026-10-12T20:29:34Z,fresh,19,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
meta/muse-wonder,Muse Wonder,Meta,closed,13,webdev-arena,Text Arena · Coding,coding,4ee1c140f9ed76faae947fb8331faa1963da41aab45ca7ad601fd698303b2011,b5ca0d3f0159af1f75ad40f8ae2c99af03a43b276ad655de3abe6118213c045e,978b3ca922cccb2ddd3c0ba3b0bba19dca3e2e258f30fef3bcce5d6aeae68ce3,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1652.0939667524244,1652,exact,canonical-source,profile_4359bef657d5e96da8d9989d,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
meta/muse-wonder,Muse Wonder,Meta,closed,13,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,49e7d98918457ebd1b6b2f3b36b0b2460b4e2a58e604165887857b5729f9c73d,2316af2dd9311e2e36a7d1a118e118732853bc97e67cddbc5a2923d0d03a7efe,0bc2bb5fde24128d8a33dfdbbf88832b7268022304f19f398c78d898747c2a2e,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,49.495,49.5,exact,canonical-source,profile_4359bef657d5e96da8d9989d,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
meta/muse-wonder,Muse Wonder,Meta,closed,13,arena-agent,LMArena Agent Arena,multi-turn-agent,2489257bb7a1ac7778c6f4d46042be7243d110b5562d714c38e80ad3b1ceb941,c78ccee7ff998dcf8c315354fffe4940105a07b899de0c0c57ef04d4e577a06c,d34b95863d5672df48f1538ac11194e4d2f59811c4a94883d50e490b03c0bfec,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.031375825809846235,0.0314,exact,source-specific,profile_4359bef657d5e96da8d9989d,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
meta/muse-wonder,Muse Wonder,Meta,closed,13,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,3ee6bf2a99f162e79321521886404b84dacce2482c8ed0f5f55f79a98341f956,c646bcf5fb4f00b9ef330e2658b3dfc7c2790ebdead592c1f92ba3b8b0239157,19b141c1dc5009047d5237719a9dc09925c296b8d6a57757b93b7e35cd947e04,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1770.34,1770,exact,canonical-source,profile_4359bef657d5e96da8d9989d,max,max,true,fixed_effort,,false,2026-10-12T21:04:58Z,fresh,19,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
meta/muse-wonder,Muse Wonder,Meta,closed,13,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,02bc09b7945ef8e709cb7297e1bf47767731cdb1e2d6fd70d7a50d1f38358d86,ffdadb36c5b86e6861a164460308b1454d858a6ce624d232f07de91ec7c20494,4007d1fa997e3bb695662565e32a1fe821ed4cfbcc00f626d79a347618190c34,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,22.857,22.9,exact,canonical-source,profile_4359bef657d5e96da8d9989d,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,epoch-eci,Epoch Capabilities Index,capability-index,0a360f30590b56a537744a3eebfe320071098ee26262c65720d3d94c53168cf2,2e08dd2a2ca2feef772148e6981a8f9e257c0cc3833179cb2b60af334385dc5f,4a8380ec7d5f2c3165205da89b379f76314783109d746bee1b4ad94d43886603,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,160.37,160.37,exact,canonical-source,profile_e90398154612163000b79efe,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,4a0aa07560ffcaa24517417d7380ef84f5b9af516cd08653c633b2078b4519d6,9497602a2c9281f7550f69a15c4d5b882f2ffa183649875584d7cd65b57546af,b725bcb4ff2f202eaaa00fa1772c63c4b3d4535616d9e0a2e420d3861c523583,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,36.234544,36.2,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,arena-text,Arena Text,preference,f6d1287f187c24a09a953df30c2879d792edbeaf3d760c1eea11192e4ed7c771,404e1d69c390082a5c0b87adf660a1919a25ced951bdbbaf9764a7bd1756393a,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1424.8248277757675,1425,exact,canonical-source,profile_402928bff9da8bc5523c7b2e,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,livebench,LiveBench,general,ec994dfeee2dbab29e5b30dd520b30e62c934753564888a2dded004b3bcf1f30,1ab80e6124d661966fb7aea3b148a2afbf17abe396b0fead25beff9851a320c8,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,67.88196428571428,67.88,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-08T22:23:18Z,fresh,23,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,d9e58069d92432a6a06fb790daddbe0651ecc36eab48935f216aa6a290281c52,84dc1ae7c2b51c381bc11d81afdd4a627a15a7e4df3a17f1b08060778a071ec8,6c04f3736f65658b96a037479925b1f1ddd9ebd0ca754f53413619a39cf552b9,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,30.12,30.12,exact,canonical-source,profile_c017fe0d6e69d734d6826051,reported,unknown,,unknown,,false,2026-10-08T16:52:01Z,fresh,23,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,70a2d25e028a6787caff00fdab2fd3909f0b7ae37b9ea92253736d57e1531d8d,1ea311d574dc0803277c38aa9255524a2504ffb90ce101ddd9acd904fa03bc3c,339c843e337f861d6acbf46534fb66c8aaa9eb4c78cfb94a9e53dca73ceca25f,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,44.392956,44.4,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-08T22:23:24Z,fresh,23,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,simpleqa-verified,SimpleQA Verified,factuality,54eca07622dd36f3338cece9c3943ab132bdd3fa7e997b98595d5a0af378d95e,882a248a94e0b62af2a5c51c536acc6e15091d94d597f98b484b25240cdfb487,d8100fd039c486c8b38dae94356690651e85c35b65a2f4a1ff005468fb0404f1,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,23.799999999999997,23.8,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-09T14:36:06.000Z,fresh,22,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,gpqa-diamond,GPQA Diamond,science,351542646899ce1e2d693c2c25eb5e728d5988b90bacd3b1eac7359fa68e24e7,9f60618b9c592c6f87b002736876027d73ec1216ee471d469b57a55ccc86dffe,a0b068f212b619b912fcd634570206c0c70e07fb8b601064a73f01a5113fe8d4,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,89.64646464646465,89.6,other-supplemental,supplemental-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-09T14:35:53.000Z,fresh,22,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,551dca864acade72f6ff10891bbcd432dc343e560a865bdedeaa935a73953da4,26be6ab66927465e56828cfdd1a37a4fe3e7612ee6d34c7a3ab892b752ecb908,492dda5a7ecb65f146bb09a90ff176ad5eb3914474f71387c7bb8b199ae656a6,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,98.88888888888889,98.9,exact,canonical-source,profile_64a24f9630001ee70ab87c45,xhigh,xhigh,true,fixed_effort,,false,2026-10-07T21:16:36.000Z,fresh,24,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,frontiermath,FrontierMath T1–3,mathematics,61ac0472103c9dc9dcbdbff8d0ed91aabfa247855664587a5ec0d90257057ac7,2962c68b70932d3a77a6189ab5b600c737e3cca12c30d1ffaa53b598682f7a8c,47e7ed13b6c1beb3d2911bc3152659a4ee1475cb0467bd3f67eaf7625cda081d,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,75.08771929824562,75.1,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-07T20:42:31.000Z,fresh,24,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,arc-agi-2,ARC-AGI-2,abstraction,9a168b37fc44e5869b3d33370b9b6a70e38e19bc02ffd7ab14207c02f01b83c7,0ef6d53901f2b4eed4b99fe3d4906ab767a18d47ee6c994519efb430e93186d5,f33ddd1da76c8fc33ef95d5e97f345c654d661ad6d878f4d8e683aa33639030d,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,61.25000000000001,61,exact,canonical-source,profile_64a24f9630001ee70ab87c45,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T16:32:28Z,fresh,17,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,critpt,CritPt,science,c112684679902abe6700b9338d34d0dc04524cd5367b7944fbca1f3f12d6c23f,651de07ae726da8d6059c5d9b6dbcf4816a739638b0a4dc9ea8a3b7f81a04ccd,7d50bb29a34aa0dd3ddd8f042157941fe5d7ec6c855d56423cfcba0dc5753b35,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,18.8571428571429,18.9,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,chess-puzzles,Chess Puzzles,games,7c18358b32ec6dabba3cd0641ec159a3f4c5d9c5304a1aea18986303a22e552f,53154b250a69a4e43d4a0b6b04ff4d2729d6d907137278b5f602bc0daa7fa5f9,f990cb74f11747d21c09ac6d5eddd2a2f4c83eee32bdb2d21a3c182c304c0299,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,55.00000000000001,55.0,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-13T12:40:01.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,arc-agi-1,ARC-AGI-1,abstraction,9c02e193d52584a567e53e0cd184f56bfc48c5796c526794f5858fe94ae4e79c,da1e0a57da033c0f9c4b03c6708c62783bc693005472d0044f0753771f6365eb,36ed7e927d9a9b2b486db826807d1ee4eaac02a7d4d493866f2173d19e6e9f3e,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,89.0,89.0,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-19T22:24:17Z,fresh,12,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,b87a58e6ea652bc9acdb7028a7289aac54b8e1562fa1a6ce44567c4a3164bb5f,50aafcb34d251fd0508e6f00bcf06139d448e3887fbaeda00f2738833151a273,a32ddabab6407a55b8856a9ec85225321a3bb2b60b8aca0feb130551845e411e,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,87.23,87.2,exact,canonical-source,profile_402928bff9da8bc5523c7b2e,high,high,true,fixed_effort,,false,2026-10-08T20:32:00Z,fresh,23,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,cursorbench,CursorBench,coding-agent,7caf869f92fa2b740490c8dd49a50a3e78fa42e893979201edd4cc2c33c6db05,4440d1988fef77b5ed25daa6fef3c5500094e20f3570aa0ab463e1f6c04f4d8e,9b0285d6898de37b7276a9b61e639721b2c47b9ae9296bca17e79fc986b53a15,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,48.4,48.4,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-14T07:48:11Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,frontiercode,FrontierCode 1.1 · Main,coding-agent,0eda78f3a0a6807b6329b58a340a555f81b9fc6b9bced66e6faa6a13322a514c,687f789a507a6d5d7b6d46f020d8c0df213395369ce13f9b77119d065d4826f6,7a7737f1d2a7a011c439011d0b6bdbe2746b86870cbb9fe1c87b5acb6f25dfdb,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,46.36,46.4,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-14T07:48:11Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,ioi,IOI,coding,1fff8429f78eab2e5633df20b76106e53696af1e2f16153c0bacde3c644e8d2e,98176b1f633a0d00c7547146de1e8130866cb50630bfc4f2fc3dad3ac7a75be8,e900ba02b462355fbc467d91ecfb5785d4b934107e6b0054cc0af911604410e6,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,47.278,47.3,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,scicode,SciCode,coding,e77722d7abfdcad300f24b088469de04b5cac7d6fd7bb9256c184e2855ca98c4,a997ca1717e8d28901b92334a1ef9dd30160d956b1baa723ebf0724dfce5a35a,40f49ec10c9af7781950ba9ed4411d7012b1011a1d1ceeb997f7f087ca0cf9c6,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,54.976851851851805,55.0,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,ale-bench,ALE-Bench,coding,8d5c3d818a6deb9dbb4a4ed0f60194c803f5673990a99a47aba4a5fb10fffdda,2b99c96dad9076be3ae8d32909c96d82b8d6a52ca7ac4f42c8c5779b8ba63448,e625ed2602378778bb8fc022e40168ce9c5bd5f85c71e4c1b1d2ba8cd1f8da33,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1587.0,1587.00,exact,canonical-source,profile_402928bff9da8bc5523c7b2e,high,high,true,fixed_effort,,false,2026-10-28T21:25:04Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,vibe-code-bench,Vibe Code Bench,coding-agent,3224215fc728b3366c2eebe655fd910f3242305edcf8bc3de561e3f338c4e8f1,9fe33981c1c9767b951c901fe38b11aa2b4b1dc77a39a4c7c55dd184e20710ff,e7f5e848e8bc62dcb56ca1b8e76c3211623f02b594c52bc75b18f8761884ad0e,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,90.444,90.4,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,webdev-arena,Text Arena · Coding,coding,2620d6e4e32f696bedec5d1c9278d0fa0f19d2af4976fa134a8898bdf8139ed6,31f893506123435fd9076784e1ca8b7510bbeba14e478e23fe13a5b7037f68ea,8c6c77c90e5bdac298b066fef6c0218a451a101aec608ef15500926efbb92b5d,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1588.8330617127986,1589,exact,canonical-source,profile_402928bff9da8bc5523c7b2e,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,mirrorcode,MirrorCode,coding-agent,237a0f6df7bef2abe510d61528098dd9b59916df4ec189ce5212ff5be59cc807,2f92168d34f887fc8d1187a32b0de3767e4a187175a2b16232c5e388b834ddcb,cb7f07777ed38d8bd52a692364ded762e8514350c8a0129c6d60d6301b709d57,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,78.88888888888889,78.9,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-21T16:48:32.000Z,fresh,10,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,programbench,ProgramBench,coding-agent,5362e7200d082540b4829af53b99d159358bffbcf573cc502a736feda6d9af3b,7be31016ff18ee92620243cb2e419d6af660653ab32c82909a6139306c33ddc5,544dce0a787479119b554dbdbbd3fd6de38f0d8479eedcaf86707bb0f20504b2,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,2.5,2.5,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,weirdml-v3,WeirdML v3,coding-agent,c10c11ade1461afd8c8346a528985d42772b7a37064b7a84c1351a3247a8a8c2,77d913d9b9f3d262874dbd930cc96bace8d355c45810e7e5071c55956a6da91b,a151f5b307e508323b3ca48d3d0d5581139f4b9b249ded3e7bee3dc003fd1fca,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,11.695232727272726,11.7,exact,canonical-source,profile_64a24f9630001ee70ab87c45,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,75386ceaad5aaf5ffbb9d1c1a44ba9665fca6be3cab8afd2342392e92514476b,3fe683c472327c9ba16782ad295a434afdf9b47dd6f940707c40145aca97dc0c,88176da786d49196c1f6b6ea2ff59bb6fb666517104e92fe7bf2aa1d690cb567,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,36.364,36.4,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,sre-bench,SRE Bench,terminal-agent,b6d637b920c0b64adb420cc6a8a6b8376b7f1b7bbc86023283dc2c6d1b5241fd,c5e94725b93009ccda507e72382aa931832a6928247b7d947844c1cc8c202fe6,629ba584332e58c7a68571bf59ab81341ddbba637b917594e714206b6ed94bd3,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,5.725,5.7,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,apex-agents,APEX-Agents,tool-use,4393651d280cb666b261c3153a14161fd4f1cb4fd1c736e7cacce3ed380f8430,ee6c8342472ac216b052a0f59ba4b8664c2b55448f8154538db9f4b2b2b395f6,a0e29de3218632f99454f1b32b0610157f76f290ecd1d47b0e860a480d4d82c5,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,56.10000000000001,56.1,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,arena-agent,LMArena Agent Arena,multi-turn-agent,e8e4efa5a1f39bd7e22917fa44c6651e09bcfd3ff06210ca8f9f3ece2076b9be,8b0aa0861d787adc3ccb3bf6840d7d21568f4e248a7b20ec18f0725c4f6fede1,33c84bbaea8304aa237c754cd12a152eb4ba4c8a7e6f92cad76f8fe65b59ddf5,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.014326641851425207,0.0143,exact,source-specific,profile_402928bff9da8bc5523c7b2e,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,vending-bench-2,Vending-Bench 2,multi-turn-agent,f20bdfeea0aa85b5d1764e324170fd6da5c70b55be1b1ff0a9d970fb4b0ab367,36da00eaff1a858a283693bf2a60d409a3c736f7b94278ab29d895018033679f,af1ef0b223163ca0a5fd70f8980fbe44d476eb17eeedd0fa7f09cb11e4dbb98a,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,3602.4633333333345,3602.46,exact,canonical-source,profile_438fca3d002dbd040e5f517a,reported,unknown,,unknown,,false,2026-10-14T07:48:15Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,68e64c18cc77361be8754b089a5dd4523dd19cebb45b10290207a7aa0c038dcd,e7ffa98945e5740027546c2ef1147da56ab85f9d43ec59c3130827d8788eb0d0,bd3849ed7ca5e0f28f15d270073add330c8d3f617829221d2fec150e85895c28,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1630.83,1631,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
anthropic/claude-haiku-5-5,Claude Haiku 5.5,Anthropic,closed,14,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,cecbb255259563978c76a4df2e75e6fb3dfa7fdd3efb61785eed7d0aec72d58e,95293cf774a399883712fc87af7112213c7d36b45ffc49707a4e5c40506d1452,3d9de72e09b5749296b8c0a74dcd519fc7d69400ba5a29d307793a47490250e6,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,10.0,10.0,exact,canonical-source,profile_1c08bb4ba5da9247e5d7c416,max,max,true,fixed_effort,,false,2026-10-08,fresh,24,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,epoch-eci,Epoch Capabilities Index,capability-index,6a1e0dd6dac98e416151d2058749d2ef21dac5ed2b4ea773431db22402ea3814,73da35b7f86f425390aef95819c29190c03d0870e0147b654f0d9630aa529296,84003d4daf797c6800487df93e17a6dcbacda825601a0581ea5b6e9455a002c7,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,160.09,160.09,exact,canonical-source,profile_5e3ecc9092fbf935153dfa39,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,arena-text,Arena Text,preference,6118c9ffc639c2d04b2a784051530adb44f9183cec069c46893069cb2c44a0a1,b2f191a55d90decb22e1bf869f39a95ab59d8cfae73818c526c810ecef9eaa9e,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1467.337294227145,1467,exact,canonical-source,profile_10b9eba1355d3375994a7c40,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,livebench,LiveBench,general,76e731809396bdb6d171e27ec2d22186babe2b10b99da863d4f8975712f85f4c,966e12ac20394b1ec008e6a94f243c9263202df50d8b207f84ac18efc9fe2e52,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,77.93625,77.94,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,b709f0f870ecf400ac4fcf0f92adf8219eeee978c5944be0281a3f457f6716be,3668031ee53d691f88bc25669ac2e8e71b7e409574de724704bed38ae05a93c2,d0afbc8f90ee9aa96eb1302c1761bc6e4863607d9c8697f73b2e6e39656c559f,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,42.910102,42.9,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,simpleqa-verified,SimpleQA Verified,factuality,215d8d2adf48ab9a2c3cda865f5475bcca7cc8c0c381c0032addb13b6473909a,43c08742fd124e14cd52e42b1f79e0ea2b62f72c136ff140da3b18aa04feab13,cf9b8ad33e4f2c0dd59dcc48ff4cf1e8b142dea609b0caac629012897ecc5c0d,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,43.2,43.2,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-08-10T23:44:21.000Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,gpqa-diamond,GPQA Diamond,science,afa0dd8c0f3d97966ca4789ba04b9ddf1402801e3e49d2031368251d06e6c467,a8b2727e7b4ae184f74a40e90aad9e73cfe2c443690c42b531e757c7757de9f1,9259e084f0476af524fbbad3e37f90d758af18a650a41c293be77a77d0ac9c1a,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,90.909,90.91,exact,canonical-source,profile_10b9eba1355d3375994a7c40,xhigh,xhigh,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,mmlu-pro,MMLU-Pro,knowledge,e2b34a3f5d5859e459ec25cb028566ae5ce9ad5b45bddee9b02469f755e6653a,fa85d1eab75ad8f26ec33bbe194d17e267c1ad1781aee9b9261d149befb3676b,f9594ee2a69f9f7a401699c6642b5cf69d6c2a1e4951ee73999b288346f6b09a,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.659,86.7,exact,canonical-source,profile_10b9eba1355d3375994a7c40,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,c3b4aec2b669450cd4942f8692c52f5df947e0ed7d44252f3c3308d65f25f348,3965510986036d214f16ea06457784f9d2e523a74ded20d8823f1ab5310c1c1b,be69afbe31d7935060cfd908634b1f0b12578c8486f2dfc26e2a3cb0cf474579,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,99.72222222222223,99.7,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-07-09T17:07:54.000Z,fresh,114,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,frontiermath,FrontierMath T1–3,mathematics,e232957c21deda367a725f0d7d3732be3465334216e69297cb2392ade7b821b4,15a006cd9b123197a9fa4fa47214c913428c159e7767398cae8e2bd36acc6e13,133ae6c889e89aa9f144f7637d831b8e5ea0fd03397f291fb8ac45efa2691f0a,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,85.96491228070175,86.0,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-07-09T15:58:30.000Z,fresh,114,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,arc-agi-2,ARC-AGI-2,abstraction,54ed6c427f02b8432ccad1d6ad55a41b9a2dda764385378f337e2b974b550651,0daa83fbf689715d2caf59721509d52fe894c8b6b27615e885472d332505ae27,5cbbd56fb7c7da82fcc952237dde82ddda0e41bb8f2be7e491076edadb0ba62a,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,83.89999999999999,84,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,frontiermath-tier-4,FrontierMath T4,mathematics,d6265695e5784bc8a7e04ece0087e3dde58d1950f531b65370a296857ad2ce84,d5eb78a15c5098c1816519b5eca5ae37f14d2ea2a0c211ccc3d9563c0d7b99b5,a24bdb9bd60f8442edd874a72bb0fb71928186fd85d815b59d6bd044216d937d,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,70.73170731707317,70.7,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-07-09T15:58:30.000Z,fresh,114,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,mmmu-pro,MMMU-Pro,multimodal,60f00c905f1c10d5db43dd22489eeb67963dbd883953bdfff309b57d9f17b5ac,2ee450b2ef9be8508b76fd52452e6321704ce41624111297ebaf8b39d659dbca,f61d7669c11176ea7984c2becaef37b06a13c7a2ea64ef1aee1bd9303c1c47a2,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,86.474,86.5,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,critpt,CritPt,science,7104bc32140dae8f0f15f4aee65e20bbf40d25234ae78adb66b1a955a6daf84b,0e4820472c1e2e83282efeb703e8a186d7cd498c82769be59c8eff30459a7287,596cc5a4ca79492f51c51847f7052c93d75b9f69afd7cf50a9a03f4d5717f77b,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,30.0,30.0,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,simplebench,SimpleBench,reasoning,6cbd9c5d671b293a6a269fa7d907d5e52a61081d799a60b893f0ba8b1d82dad3,c5664b0c2edd908342147bf6754442b4a98f946226905e3dff21716e4c147d56,902409fd7a7a4a66284db7f42458a2c436474c8784c97627642233ce61ef4bbb,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,48.9,48.9,exact,canonical-source,profile_10b9eba1355d3375994a7c40,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,chess-puzzles,Chess Puzzles,games,f9dd099bdef26913d6265320e408247103a3255cd603e07035dee8c0f32e4c89,61be4778176b234344beda928d45da0576d3c2f56253746d38eaf1aad88b20c6,f7747f43de11ae40b88991932308c7af56eb77d62c4c3bb1464a84ac8347bcde,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,47.0,47.0,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-10-13T13:53:13.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,arc-agi-1,ARC-AGI-1,abstraction,6fb2fb527bf34e516fdc871fdff2adc6a0f432cc04cf0c20624cd37d1c5089b1,f49fad63a8378f116653eb40aa4d5be3943c8bc21e15b067ccce8efe636b29eb,f49fc5d5dc17d925e46a34e90ac27d33983718b43bdeec786fc58172835b5c89,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,96.5,96.5,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,swe-bench-verified,SWE-bench Verified,coding-agent,a805497b4515cb42c17841e6fd90b7dd8ded2ad89f08cc13bcb756f397f77818,de8b52f7100b63ce8814052d4602d333bb4f5ef964ea27a537cc987ea86e4f31,9f593eff97eded368a2a5202ecb446e3a70d257e0b15e124388293ec2b52e7bb,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,95.4,95.4,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,deepswe,DeepSWE,coding-agent,f8f4813d3f8b1933d5f29a92aaf14ae4f72e9829122d8a04bd22b3d8f0d14437,bffca5bde24465d231f8a2a3813eb75aaf5da1ed2736e912ccfd7b929ccaad93,869d4cd8ddde2355153ef770f54e40b2142f9c58a3a57dcd9757271830347ed1,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,69.62305986696231,69.6,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,2f457bd2892da6a5a93023ed588b36980a4057b4887292cd4088c43c8119faed,5db62aa82836c3f8088a5b982de5f69aa07bd00fa94c7d1eedc0681765111439,9c3e7cc3281ad1d5506c3750c94646d76190650f0e457545194659f17843429a,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,92.37,92.4,exact,canonical-source,profile_10b9eba1355d3375994a7c40,xhigh,xhigh,true,fixed_effort,,false,2026-10-07T01:44:58Z,fresh,25,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,cursorbench,CursorBench,coding-agent,64103639d186cef0cd17bd10831c913c8bb472733c82b51092c3f9c011724d98,bc35420b05a54293a438b8cdb679d414117e5655cd4d076c5ec1076bfdf15588,f672612a035137f64e1339dee647878cc4bc1d74f31822e6d7df166bab580883,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,41.3,41.3,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,frontiercode,FrontierCode 1.1 · Main,coding-agent,279082650a63709bfd0578de825a500ce8db081a3f08d4843b5cf016de6b195f,cd0d08a879a68c3215c7ebd97c1779a18ec468dfe6cdc9176b9c594a4293df21,28b116966df3105e70a47466a630f92d7e1cc8b81dd0f1635b10ae1073527658,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,41.31,41.3,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,ioi,IOI,coding,e5e3bb19c310f7238c0ee6cc96812e62b3f9b04ca5cfbecddeba9c0557bfae59,0de9a663db9756407267c731f1c4b0ab93295ed1c7fbc671737d793afaa647c3,030e6f30c5b5f4c6c26d341d3d19f39f3f0c1536e78b5ce9e347ba1f12247faf,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,87.611,87.6,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-09-14T23:34:42Z,fresh,47,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,livecodebench,LiveCodeBench,coding,eda18a8481dfbfd1f072440f2a9e46d2f3aaef65c369c6012b7e98c4c3d6831d,9ba42b1eff6f928b3f6c1987ea7bb8a707b437f978714efda59c25896b4cf272,4a8010f008cb50516a1869e16177d943a175d04381a8535bfe752778d45fccc2,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,85.93,85.9,exact,canonical-source,profile_10b9eba1355d3375994a7c40,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,scicode,SciCode,coding,2077c4f18de65cc728a797e57449a6f77d5906814f5b554bb8e89b62cfdf702d,102589cbc8c681b26512485a1ad4c8d4031c5cd783b45c7041efc95ab017c857,3f3b163f8a0e9ff3ded55295ab9a2c86e218e7202c444a20602d29a924e7bc84,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,54.976851851851805,55.0,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,ale-bench,ALE-Bench,coding,4b73b41a7de5ef70a58fe2b121d713c3aa91a61e2d47d586992b68e2c5d29bcd,a7cfe357f7dc3b68018fe3548dad7cfdf12a6a8f185193ade0b1dc7f5e9d3e38,27f0be14ac4f083323244ac5b6ffd62ecd97b3194345e3ac7bd2b981343f7641,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1951.42,1951.42,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,vibe-code-bench,Vibe Code Bench,coding-agent,b082e85d4c64ac35ab1834c5888f06b3984ec9cedf4f1dedf1c79e9b6d73af1c,bca05426d3268429d7c3b87f2c9da0a8f7470fa4f47ee3f540c51126d789ff2e,71e22a4e7f483e29770b8f34273d2da8e3cf3d23bb4d9a57cd99162e7b8babbd,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,74.594,74.6,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,webdev-arena,Text Arena · Coding,coding,508235f00d0f73ddf54c782c11f3d619eedc14afd7e2a0f7ff3ec1bef76ed8b0,ec54b9d2a8d653faf2bbbd6dcc88f2ee7373c0ef143965861d0b91b7526f67d1,c783a058711fa756b0b4c0b6a819ebfd5e28e53155c8de9f59425059f536a44e,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1520.1279670236868,1520,exact,canonical-source,profile_10b9eba1355d3375994a7c40,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,mirrorcode,MirrorCode,coding-agent,af1b21692f16906a8ddb3160e6267b102e85fc851fca4c3e2e8885978d4e654a,52b6405657288db54ba42354b45554962b4dcb0940faac3b27698d655ab78c2b,509b8fb4fca90460d340b424975a2299f1777c3a2aa05ded4252d395d3e93606,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,41.11111111111111,41.1,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-10-20T20:32:26.000Z,fresh,11,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,programbench,ProgramBench,coding-agent,7f731ba0bed3e5c1929ccb28b3c1f206eb551be2fc82eac496af15325c1c8ef5,2837d20ce59ba49adcee7ddb57ee72b40dda5fe8f394cf1ad1c149e8f088dabe,785c19dfcc17a217bf43871e003523ca62db7d85c08826a350e351fd85b30997,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.5,0.5,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,83698a8a2edc3c85c45c75f86ddd31889360005d421b380baf1e5a0d5c1de779,02bb961bf99f8098067f60242182fc3c757c1902e74b7dd6e2c5f7c767a635a5,34c0fab0e6346dabe9646d8debe4dc216ea32bad33a0108c393ed5f6f43ff452,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,77.528,77.5,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,e8c624ad25eb75d29f6b71b1ad10284e821c36ad7e714874731b496030934bc3,e059c72c91c14e650a68f5cadc5059c8866692b5ad1a961fc7031be10208205b,4f9cad5114a3ed9fe0ae7bc0a047a19b833e44aa7c26716f9026b615bbfadd28,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,22.727,22.7,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,sre-bench,SRE Bench,terminal-agent,b80853a0bf0bdb75ce23d43bbd0583c12020b63673cca931ba38de7dcf42aa42,fd79bbdaa60e95caaaa307263a256d01e1f6c6e3f772a04002d108bb14123b83,4a88e5b3a014fabfb8c0519eeddb0aaad2024bb7a038a6b65a5431402cfd3458,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,9.542,9.5,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,dc119b24de16552a1bcb614dec058d988952f1a8aaff671597fa2c2cef83839f,dfc9b375db76d7b6826d46b7e1929052cc6fc2836418aa2429f26d4803a5f876,88ed41acc102afe2027af4516d6adba5f83278c801741a5692fc562408cbffb2,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,40.206186,40.2,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,apex-agents,APEX-Agents,tool-use,e750eb4b30cf5308f78fd5bbb314a752bd68f40f77528b12e5c7ec3c64cd31d1,2ea7bc0988ad760874756e1a297abffbdff13b0496ce941e12c24904734b3788,fec372c219c2731ca38a304996301cb2ba615d79fd83872b61b1f4976b94aedf,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,58.20000000000001,58.2,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,arena-agent,LMArena Agent Arena,multi-turn-agent,a0769915514f42715f1ee64c21b1f2e5b4c23747697c1cd062a08a3fb1e53d6f,41a8453a6106fac3336d2f29b9d8975977fb1c99364a4f6b10d60a0637b9daef,78ead7f38e717d142b1ab05cef4ec431b5bdc9e72e041c0cfdf17df2e50cdba1,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.013902435627719558,'-0.0139,exact,source-specific,profile_10b9eba1355d3375994a7c40,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,vending-bench-2,Vending-Bench 2,multi-turn-agent,4a58472e3738fbbbad9cf162cd1406eca1e51404733838aeed700477cc57fc3b,945bee3b92aa43c98e4059ffe61d3d6531a7228485b16053d54648e9ab7a0f07,a513e3bc1170493bbe4a4f0ea74a94f7ada17a29a737518886e56958e0c294d1,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,7343.206,7343.21,exact,canonical-source,profile_0770c98a055bcbf0b3bfe388,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,fb5bb99f4a7b99d7e8f9c37c962fc65c2ab0d1cf38d0db715384adb4737ed3e3,1ecfad4c6d855e372c5df9e4fef63bdb3bfd40e59d884b5e4344fb473e13d423,4438714d09a3be3f0ddb1e48f461efc662d3bbfc2af940f9d3784a12162e10af,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1461.46,1461,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,terminal-bench-3,Terminal-Bench 3.0,terminal-agent,c11ea58daf33dc934917dad6b9ed9c08626c6a2c192a071ed482a921a4a4d2ff,763202f791382f7b7ef5da41a61a23d7342d7f54ffa25ffabfe00a8c2e77283f,a3674846c9dcf29b2407ae67a492b1d523022ddcfd99d682b436215d9b3be416,release_ceb36f25882d8cea2c9d084a,3.0 · b936386b-4afd-4ad7-aa7f-6b52bcab9853,official-74-task-leaderboard,accuracy,percent,true,20.81,20.81,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-08-12T21:39:34.661007+00:00,fresh,80,0,A+,0,A+,0,A+,official-leaderboard,4,official-leaderboard,terminal-bench-3,Terminal-Bench 3.0,https://ofhuhcpkvzjlejydnvyd.supabase.co/functions/v1/leaderboard-read
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,947edc512a3e891792825bd53b6a3e17400a6156b27e68c7dc1f5d6ab6193137,99731683cdfb6a980bd718765c99c3ae18a73239e779c85e7c3c1ce1b275597d,f86eabd07ed6cd5c58f3bf788e838786c6ae9596806380bc0886b29d04157b44,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,10.0,10.0,exact,canonical-source,profile_7a539a480ccf940eda2cf169,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
openai/gpt-5-6-terra,GPT 5.6 Terra,OpenAI,closed,15,weirdml,WeirdML,coding,6729b7f8b7fae0aa6b463bf70f850e788b118fcd646fc5e9e1d9a9af8f653ab8,2fd388997e65bfb3841305faadd1286784bbaa658533fd6dd4c43f6f4ebc7c01,c3a637f5e3e24eea32adeca84ed74d99473b1db1a5d4d3c763e35c76d785f768,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,78.27,78.3,exact,canonical-source,profile_70294a052eedaa815d9395a8,high,high,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
openai/gpt-5-4-pro,GPT 5.4 Pro,OpenAI,closed,16,epoch-eci,Epoch Capabilities Index,capability-index,c55c62c0a20b365c872d36abef039e19cb3f2edcfd463ed2a57910a44ede2bd7,5f49abdfe7a0ba61d17467f9206f12dce6003db890a4a00ecfbc62dceb3610c1,fe3259c331bc15dac88d3c255cb9da15e63e76cd00edb96ecf08eb0651040931,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,159.81,159.81,exact,canonical-source,profile_810ad01488e903e7ff5a14ea,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-4-pro,GPT 5.4 Pro,OpenAI,closed,16,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,311fe60c11458df966a2c475ae5bd4d884ab2f8efe6c3d6571b5cfe710ec0f76,e39b68a562d118faa454689251dfa151768a19b3e8eb3ab16feaf8ac7adb8d1a,0437bb5466e0106959c9156fc6db8b14893f980280a0349157c33addcd90bfdf,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,44.32,44.32,exact,canonical-source,profile_d9016c6789c93f7997fae173,reported,unknown,,unknown,,false,2026-03-23T21:12:56Z,stale,222,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
openai/gpt-5-4-pro,GPT 5.4 Pro,OpenAI,closed,16,simpleqa-verified,SimpleQA Verified,factuality,f813720596c4d46ed80f58f44fdbb17d388a9adcfe5523b84b2b7ee330cc0c2c,92dc6ef9d2bda5b1affb64fbd256c81bee65e371c1a939445298f9e88a7e1f7f,7a9312bb814bd62a83ca8daba17fab3f15dd4cacf4f0f7d654c6220ef25f94ea,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,46.300000000000004,46.3,exact,canonical-source,profile_b5490cd7180da4cea2102329,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T19:28:24.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-5-4-pro,GPT 5.4 Pro,OpenAI,closed,16,gpqa-diamond,GPQA Diamond,science,c992b13120d5f50a6aee9a19d1bf911a395f343ddcf6fefd31af09d04a746bc0,c1173a657b58caa2004110529ecc167d421a9814bada59d97bf62eb74ed33cab,e472fbb92aa2c73008c634b78c57262b905e36573196b7808f984bba5edb23f6,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,94.6,94.6,other-supplemental,supplemental-source,profile_b5490cd7180da4cea2102329,xhigh,xhigh,true,fixed_effort,,false,2026-03-20T00:19:39.820Z,stale,226,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
openai/gpt-5-4-pro,GPT 5.4 Pro,OpenAI,closed,16,frontiermath,FrontierMath T1–3,mathematics,c35087f7442b680df7d9d73299a80ee0f187217edf3ea3b5dcb8a99952e88c85,f391a5e8ca20523d7cbdddbe05989e74fd53bc7ab2f58385b02babb940bf3feb,c8064973eb5564782e385b99cac1d2a5363d6786ade312eca4d6beca296ca1ac,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,82.45614035087719,82.5,exact,canonical-source,profile_b5490cd7180da4cea2102329,xhigh,xhigh,true,fixed_effort,,false,2026-06-13T02:38:02.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-5-4-pro,GPT 5.4 Pro,OpenAI,closed,16,arc-agi-2,ARC-AGI-2,abstraction,b7b376846be484c0de704db9e09c9a311c08c093a8c867728a25c3f55b5f463b,cd35d68abdfec229206a312905cc34c1a1692c02755d7d8e80852e5bbfb8e6c2,d70d8bdb8ded770b2df917285d64d3028140c60dbda01cec10ac66a81e6c4e43,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,83.33,83,exact,canonical-source,profile_b5490cd7180da4cea2102329,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-5-4-pro,GPT 5.4 Pro,OpenAI,closed,16,frontiermath-tier-4,FrontierMath T4,mathematics,0560402bae1ffabb3092d47e7a4864bc41f4dbcf5dd9b37f424fc4ae2d2982b6,db774535d3ad057d3a4e8ed672736934a1ffb95758e2e636f8347d9403217339,cfcfd105753964b6092a611d388de2bd1d1b6f3f4e2ad588abccb20d09c05eba,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,58.536585365853654,58.5,exact,canonical-source,profile_b5490cd7180da4cea2102329,xhigh,xhigh,true,fixed_effort,,false,2026-06-13T02:38:02.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-5-4-pro,GPT 5.4 Pro,OpenAI,closed,16,critpt,CritPt,science,d370876a81f1018ca4d35661584babac4ff139009a5b08986f00568f656b0f19,a192f0eb1c79c81f3801d96d9e6cafa09a94492ca1b6a09e56bfedb16a7315e8,20bad2d02546d0dbea2a4202e81e21d75ec48fe9047a59199fe4024c7b4f73c4,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,30.0,30.0,exact,canonical-source,profile_b5490cd7180da4cea2102329,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-5-4-pro,GPT 5.4 Pro,OpenAI,closed,16,simplebench,SimpleBench,reasoning,90706731f3eb4e40c071b93ea4ca6e4cc5ac60fa9187e234dba442f55ea5b500,c349e5c8a3423545f594717e2fcb519fece2677a0d3d25140e8989afd41b1a37,8f34d561d4c9ab49272d860affee75cf2fd058ae445783d63eb3ccb55e8e7f12,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,74.1,74.1,exact,canonical-source,profile_d919041bb5adc55760362ecf,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
openai/gpt-5-4-pro,GPT 5.4 Pro,OpenAI,closed,16,chess-puzzles,Chess Puzzles,games,1214adbd9b742874bc90b9f8c889927e741a932e17e0caf66b5903a1f972b04f,ac536d4b3e4f92b96702db2a95cab41d2d8a4eb249586934b217d37bff717d43,d473ce7b9e2743c267f016af99f527f45d075aab2a94d3fb3a9bdd16a76d5ad2,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,59.0,59.0,exact,canonical-source,profile_b5490cd7180da4cea2102329,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T09:56:24.000Z,fresh,17,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-5-4-pro,GPT 5.4 Pro,OpenAI,closed,16,arc-agi-1,ARC-AGI-1,abstraction,aee2ce79c54f08fe76385d4cd5f478e829b4d770e51c98ea1681cc588413b6a9,40dffe2095bb778fe6d51ddea6fce7bf3ee5f752f0dfc13f0e79122d405c6bcf,d3620fdda0fa9df1c5a9409fcc1a63d89656463398c9dd19a03fe44da9701675,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,94.5,94.5,exact,canonical-source,profile_b5490cd7180da4cea2102329,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-5-4-pro,GPT 5.4 Pro,OpenAI,closed,16,weirdml,WeirdML,coding,7615cfbcfaa999bf4f630ae12cf22ae8f7bc25d3eb7e276efab0a7372aa2362f,3f8d18e1c08c68151e6e04c047208e7bbbfaafac41daed1a1b349b9c4318f056,24f3d247a66f5abc9d7b5ec648f3fddb530e8917ded832cce96c305b839ac051,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,57.440000000000005,57.4,exact,canonical-source,profile_c2e76640622cf0a45f5bec51,off,off,false,none,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,epoch-eci,Epoch Capabilities Index,capability-index,69895035dc168f7780536eb41da2a428d7d7dce11646a50866ef06efdab51684,55de49a29ac189930965ecb61996445143ab9778b97d8ca00413c28cd262789a,ef857165a5b10fdf19e8f420846f2ccb73e01232080de55cadbe4728731a2864,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,159.57,159.57,exact,canonical-source,profile_19fe5fa33d354d1d1a022388,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,arena-text,Arena Text,preference,7090728c577fc41d1f68ec4f32641fa8a52db2e9ef03aadc3d9d5f0431c3929f,12cb5102272989009a293acf4a7feef24e6863d034e94407e04ffb263de5c763,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1481.711070665865,1482,exact,canonical-source,profile_3ff20fb03ea5572c13bd7021,high,high,true,fixed_effort,,true,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,livebench,LiveBench,general,6f6b5c8b08b0d34db071cd7c19ab25c8b56cc849caceb5b8011b23e32f30f485,06c74994d7ac87c262cbe9069939ba11e776f3b7f4c01dcc082d392c17edfb68,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,76.22398809523808,76.22,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,b082396918f198ce2501b24ea8e2b8bad459f88e9f6aaeba9513c30b48eef83d,1810b710761c27cc29697c1b0fe2e6067b5a83a11a8f3b4241f2c3a9be7d4859,8828394c5ab348b4e326438defb67fbb61251e8435b7c24d33f97db6f8d78421,release_c1afb6c54b8ceec8bc1f121c,HLE original full multimodal,full-multimodal-2500-no-tools,accuracy,percent,true,49.8,49.8,other-supplemental,supplemental-source,profile_0dc76ca4c859b3a6a92b7b6a,max,max,true,adaptive,,false,2026-05-28,fresh,157,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,anthropic-opus-4-8-hle,Anthropic · Claude Opus 4.8 System Card,https://www-cdn.anthropic.com/0f0c97ad20d8005706296bd92aa1c27c6b2f4f61/Claude%20Opus%204.8%20System%20Card.pdf
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,simpleqa-verified,SimpleQA Verified,factuality,fcdf130f65dbfe4804ddbe94114d561e3184312563d22b1d716492a37a1dc44d,1d93287d2a8d4ea0a3bd161f60b9f2fc77d8345f4f0592f84fb902a30f8db5e2,8048e6f0aa21e53a5260b9c83d5b08a693fa97069d9c4dc25a37482175a1a6c8,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,53.0,53.0,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-08-27T19:27:24.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,gpqa-diamond,GPQA Diamond,science,f9c1ccb88ab1f56f79bbfe79e72186b376bfc9c90cf68fd1b040944f71063697,82e901da5f5f44589f29d61b7e4daba50f8cfb6ceeef754a6faa6f3438019f22,a7bcf8a4474e072f185140710fbcbe3fd0978faa8934b4e9f23be6351419ad5f,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,92.424,92.42,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,mmlu-pro,MMLU-Pro,knowledge,1265f4a911420a92004ade350120b568271069f00b09e6e430573e873572762e,9afe67485f8eac7e03485304475c55d3bcc960506514cf2933ef79b68c59029d,854d656fd3e12bfacdce72437f31b0d01fd553ac2e0a5e999e664334be5d0d96,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,89.585,89.6,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,174463ad9672032b5f31c59d9d2e7a35b92c14b204623de7cb6360b6c4602abd,8e853413198d547a958f7b19848455cb9a100abe30a078242496951c835aea32,3363252c17a2fc64205757514a0be605af0eedfdaae23c490b34e31ca4327790,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,98.33333333333333,98.3,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-06-07T23:44:54.000Z,fresh,146,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,frontiermath,FrontierMath T1–3,mathematics,fd77c67e665368bab7981e3bfd6180a1ad39a4bb77c1c7e14d43bb3572dd8a46,bc6862118a68f29930153141c3eff689211765cd89d0ecfcae6e9341c88bbdcd,dd340eb941ab0c21506a3592392589d334adab0312a07863d24010721c5eb092,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,80.0,80.0,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-06-10T15:25:15.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,arc-agi-2,ARC-AGI-2,abstraction,3da4350f9c010ec86ee5d8b00660dccfcb9965a7d1e6dff7a532136566819bc0,72da4da2a24c118696e25d0877da5582ed1ca9cc9bc70ec67e229cd766bc309a,785a72df191545092d039dd57501ca156b8dd8150479806c6b50c0373c9796f5,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,72.08,72,exact,canonical-source,profile_3ff20fb03ea5572c13bd7021,high,high,true,fixed_effort,,true,2026-09-02T13:33:02Z,fresh,59,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,frontiermath-tier-4,FrontierMath T4,mathematics,2358d47cac15e978cbb4c0b4ae399d02ed8c4d45583ab59ca75ffe1862f8d89d,d3dba3667d637752967aaa8ba99e11dfe4682903743dd8aed384e40d517e0b8c,75931d8996e8e66e261aebde906fd9ff2c0c295340bc18b20aa6eb2f379cba06,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,56.09756097560976,56.1,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-06-10T15:25:15.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,mmmu-pro,MMMU-Pro,multimodal,1edc74a3fd36a088835ece6cacdc346d7581d5c06dc419b46055175dfae950c3,6fc3d2bbd3d56f55222eb327222bba3524488619796ee800898f6ea397641406,d91ce15eb909f431a19992cb09c1aca45a1aa8bb3c776f678f7b31376fc676ef,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,86.59,86.6,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,critpt,CritPt,science,ee2ca80d8e2b18dc16ae537aa2f6216379eeac74f2547cd4b19211ed6174aac9,2efe2ad3f337edb173ed3845366befe82319ae9f180350ee4e62c9335b76b4d3,eae1f751d22a1666394481c64aad373db1009d5cf59454f549cbd63424bacc61,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,20.8571428571429,20.9,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,simplebench,SimpleBench,reasoning,0597d1256ac942b4d5e3d6a3e649b2d08ad29835101c1c7d0aaa23a2c84aa15d,3e2237ce4ca28649a34b3d7318591fd60cc2baca23c069f273c4eae828470997,b04a4db496b6f73099c0c613de0dc31bbeac6138402dfe202360cdaa95e7907f,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,64.8,64.8,exact,canonical-source,profile_c95142f46157a2f984bb1f9f,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,chess-puzzles,Chess Puzzles,games,18227964cda56c962b6c7c48015e51c13572f13d8936fe3d669a2043cb1d3919,9249b518702e1626a876e9e0319acd834520d0f8b82ebc718d8f7a6f30afce14,62b52b71b2ccb9831d5c233c3ed19a3604763b637badb1f1ea577a9bcc05a7c4,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,52.0,52.0,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-10-13T13:35:27.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,arc-agi-1,ARC-AGI-1,abstraction,63508e0885fa54fc627b672aa7d56803cc249de871ddf196f4367c92c3f8b12d,119ada31b69c498de358109a11dc3d31e6bbe0cce7f9adffeb5b99ffae395423,9f87e4176f2a64b2f0804476be952e926e8821a4e85daebece5d9588a07bbac2,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,92.5,92.5,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,swe-bench-verified,SWE-bench Verified,coding-agent,076cacf0eb09b44a9123b7c22dd131949bc6d3cf4de250bb105c3673ef1afbe6,e365cd1cdd8537e801d36c5b63c385e96ae7b7158302e6a1813b5f2594e58993,25947717f47e88fe72a2835cffd7b891c770f077725801a16e67c7e2ae03c81c,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,88.6,88.6,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-05-28,fresh,157,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,swe-rebench-v2,SWE-rebench v2,coding-agent,dd2c952d23d70dae7e0d09e246c2bf130c357575d42d7ed66a994a59323c60f1,365f38b1eba285d4751b16405bf0980756ac038a22f444a4db94202c26c59ab4,7645eeaa8bc0d9f452dcbd478ae9fc099df4ca50dd375c7cc8a3963a612ee618,release_72d764965dd7baf34f983249,rolling-v2,2026-03-01/2026-05-15,resolved_rate,percent,true,56.545454545454554,56.55,other-supplemental,supplemental-prior-window,profile_576644939af96eac30f5bf8d,xhigh,xhigh,true,fixed_effort,,false,2026-08-30T11:02:42Z,fresh,62,0,A+,0,A+,7,C,potential-contamination,6,potential-contamination,swe-rebench,SWE-rebench,https://swe-rebench.com/
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,swe-bench-live,SWE-bench-Live · Lite,coding-agent,bb0c285d91c12d6a1a2964b3ed76a42128fa8d4b153eba05607076b800c1540a,ac608d2eab915d6c9cdb6879bbf08841602bbdb151cf143585908aac24b424b4,fbdc90daa6cf4b367e872ee26db4a4f860d827ec2230943616878e70af968b92,release_9d2de9d5be7e9822af19f7b3,SWE-bench-Live Lite,lite-300-python,resolved_rate,percent,true,34.0,34.0,exact,canonical-source,profile_576644939af96eac30f5bf8d,xhigh,xhigh,true,fixed_effort,,false,2026-09-09,fresh,53,1,A,1,A,0,A+,official-submission,3,official-submission,swe-bench-live,SWE-bench-Live · Lite,https://swe-bench-live.github.io/reports-0605.jsonl
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,deepswe,DeepSWE,coding-agent,1d530abfe9a116a88ad444aef13b9e074953d117c33e68d3e1cf2e7d2e1925dc,71a0f61a00fa655e502a3f9271ff3bd4787f6ad87f5eac0047c2db9393e1ecbe,4ad987fa0c4aefb8ce2c11684129ff04829491fd3bdfed633b66139aa56c3c1d,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,54.36241610738255,54.4,exact,canonical-source,profile_576644939af96eac30f5bf8d,xhigh,xhigh,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,gso,GSO,coding-agent,c533cbdd1802c82945b2e917ed27a0ddf31a898791e4e052cc3c7c405b054199,25ca9010468707815cc61b9dca5b279c496c4b6f0089a5240e2ca621679c0713,d3cca460ab4bb7193a78c5fe79da02aa8e3f4b1608bc2f2d43a1e4e0f62c44dd,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,47.06,47.1,exact,canonical-source,profile_362ecb5ef36351fe2feb7757,reported,unknown,,unknown,,false,2026-07-12,fresh,112,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,frontiercode,FrontierCode 1.1 · Main,coding-agent,9049f8d808935c905e5ecc78f1794c97f1da290a808ca9b97a1b381e3c7a512b,70d55f2a031712951adea002f68246f5071df74457fd9e257f399c9c4f974492,6c43f8697f1e896c86b18155ebe54af2143529dfe05f19c2b87a6469d8be96e5,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,46.5,46.5,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,livecodebench,LiveCodeBench,coding,ec34b020d99fdb58074f33f0c56fb1f6ab98d103b461a5781ef57460b9754a25,4c42ad29713caf2f52aff64c63daff2cf98bab5639399cc89ba17b75bdc018e4,7d4e4718d7dd970dc7838f53b8895871966c9f3ab0380b65ee7707fbee42e207,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,87.819,87.8,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,scicode,SciCode,coding,d12bcfabd882e907c7f02ad464e5cc20ad3feb142c43b1139de6174277dc61d9,d860d3ec9e8698becc4e4fc17ce36713c1ad5f910ede779a81110ceecb98708c,7fbb68e2e35d45d01ecc52548723a571e74566f1f545239950b3e54ac27baf7c,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,53.4722222222222,53.5,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,ale-bench,ALE-Bench,coding,42bce2e8ff4341440c29f00235325f3ce8f804648e2bb629c670675b4edca37b,d594c8cb91ee07449d415d8782e5af74cc622af414264a97cbd1d2bfa1fc474a,2d92da04a443fbc506c1f2abe595602cd41f70cf835a170a1b89821baaa2b939,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1563.83,1563.83,exact,canonical-source,profile_3ff20fb03ea5572c13bd7021,high,high,true,fixed_effort,,true,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,vibe-code-bench,Vibe Code Bench,coding-agent,56b8b87bbab31e3cbd3fcdf78218386dbd8975b43d8001cfab1b14faf3d10bca,fff7833d3d35cae67882b1ba189089567275d2950c4a1b3160428a2841881f06,2964986b13c320c9b670a439f7c5e105bd709a89ae2ce5638b1afb5ee977b415,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,82.725,82.7,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-05-28,fresh,157,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,webdev-arena,Text Arena · Coding,coding,a349b460cd9a2b0de8c1c2556184a4d00bd4f2afcd181a2257f0e68b68de3b7f,dd72e52649bad0e0a733c55fc0f0aa64f350d72d590c688f21e87e7584010451,9c7124e258e9e5a9a057179259a0297da1455623e3baf5981da9cf83a005a98f,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1555.8263524171814,1556,exact,canonical-source,profile_3ff20fb03ea5572c13bd7021,high,high,true,fixed_effort,,true,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,mirrorcode,MirrorCode,coding-agent,584cc225fd4684ac6f91fcbaa95c681ca07df1ec1765e0b440a56781c6576830,50dd3fec98e79d70524564314acaea22d2b5ce4701aa433b3908b1f7f28273fa,c340eaa940cfc8d678f19af1207e8939a829e75a57551a135896eb037dd7b385,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,65.55555555555556,65.6,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-10-21T16:46:28.000Z,fresh,10,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,programbench,ProgramBench,coding-agent,9de7c361a04b16754281e782eeaea14055e1a1d2428c05ed30227d5bb8d67a72,4fbf34477f4ba03f9567d031f27232df00fde9914e3f776b56508ee702f21215,e69dd949165c1c0d6908f5c40d22d7980598c03636851a95c34a00001aca7885,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,1.0,1.0,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,6f01914f4b4532522b4fd3c491db69c13b9427b077d27b32b9ae09afd5deaba0,3b654ebcf8acc03ce0bb4d8a10b6715bed6742f3627c91b273a3e9d35e01d447,e5c26a91bc4de32b3edfa3a48708711fbb61b5e75399afc197e3c51a243b8020,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,71.91,71.9,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,be2682bc268df5801681287ff800a37fb389f8c742981033e2e97cd43da63e95,6776ab7f08e3f577fa5b97d73d171aa3785ff3c989f9455c60ca90328e40be59,30cbc98f901ff15467520437bc3cf91d0b16481537f4a2d26940681f7d7f2756,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,23.232,23.2,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,mcp-atlas,MCP-Atlas · Scale,tool-use,a2d793d9241d170a17e280fd6a1f4489b2108cfac972d69d00c2fe53affb0591,599ff29ba2908d81880f6bc5945a60a6dda8fdaea520711815ae770216e751e7,0a9bc5731a460d066225ffbac647448f1f8405a37f07f38637268204fa017769,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,82.2,82.2,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-05-28T17:37:01Z,fresh,156,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,skillsbench,SkillsBench v1.1,tool-use,967d0832a5742ff76f01e9f8f0cbc3a53196921113e7f79e147b0bc88db41749,b2d7a9c072e3c4e8048aa076970fc9face7a6bc18ddb0b75b2ac35eb8d095281,c2ed2025e534826e5e627446f4493a128269710491b35eb2122abbaf19d56304,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,54.1,54.1,exact,canonical-source,profile_63ff087821172cc7e101cfa5,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,apex-agents,APEX-Agents,tool-use,ab4098b38f43b4ebb93a81c10ed38cccc325c93b20220d28f1fcb19c3a23c8ec,6aebcb806ebadbbd98711f5228fc40f227f2525111b0474d5e74f6190250098b,82ea4a6880dec1d92a79a955a76dbcc1519cbe28021f6ca3d4dd37556cbfa3ce,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,48.9,48.9,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,vending-bench-2,Vending-Bench 2,multi-turn-agent,57ea11f4c2545983ab7ebbedcf108c3839dc3937c737fbf74e6a5b51a4911995,a59dcc68fef1e04939e7947d7e72b68c3b431673c9a74a3c065c75c0a2529edd,c425bb83180b8a8f09c0a10f808a35acd64e6fafd4b7a0c98900a56b812fe5f8,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,2992.344,2992.34,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,e6e2f30699dc1ade952251e4409a65fb163f3324da49ee35d48f369ead9cd41e,f5c5105eaa4dfe0637a7dccfec7e22f5ace8a85e0737022b78a2cc63fadfe55d,1fa926baa74e6bdba5b0b1b08401bbd351b83d6a5aafdfe20fb555e451bdf8fc,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1461.75,1462,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,terminal-bench-3,Terminal-Bench 3.0,terminal-agent,687e9341980bd7dc7c3fb4b77dd726c93c64312a30a806f3b4271c74369b08aa,523813b6acb7d59ae02893f63a8cb297f306a5666fb62934eade13ca89100122,88160db45de7ebcb4aa5f678c797f77980ae783f5fe91a0902414c7a82885016,release_ceb36f25882d8cea2c9d084a,3.0 · b936386b-4afd-4ad7-aa7f-6b52bcab9853,official-74-task-leaderboard,accuracy,percent,true,21.08,21.08,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-08-12T21:39:34.661007+00:00,fresh,80,0,A+,0,A+,0,A+,official-leaderboard,4,official-leaderboard,terminal-bench-3,Terminal-Bench 3.0,https://ofhuhcpkvzjlejydnvyd.supabase.co/functions/v1/leaderboard-read
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,5a038187f3436b72b8687a4ccd2646e94a883fa3e64e2cb51ea9aeeca79480dd,20ac6cef8372c2247e5559da51efd82f253f35c58b0904246ae9cec370cadd5c,575a6455222e836ab6aa781653d058f696cbe35556cd834d0beff1868bff9a79,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,4.286,4.3,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,osworld-2,OSWorld 2.0,computer-use-agent,e1c4905ec5b87860e17d024ff42801501661ce08dd8e4506cdc521a663770e1b,1f8bb98d0560eca1726e0f795c50e7abf586cb96d867aec962499e8f768eb6a2,3f78a90e46b6e5d1e3b3d6dac19d9b6b42d15334f71d453e4d086ee22bb1f582,release_15c5525fe39f33bc181dfb69,OSWorld 2.0 current leaderboard,binary-accuracy-500-step-budget,binary_accuracy,percent,true,20.6,20.6,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-09-02T09:27:06Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-osworld-2,Epoch AI mirror · OSWorld 2.0,https://epoch.ai/benchmarks/osworld-2
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,557ae3b97c9c7457b034d32b336317bfd8d10b9dc8c9fbc55753a44925156c22,29d3083f87533eb903cd2ae19e9fab477b18d93d30dbdaf0a2b7e5c8d49595b2,88e9ba377425eca2d26ca6bb75fce7b15a85a4bada5fe73001024ea97041a534,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,39.69072164948454,39.7,exact,canonical-source,profile_5ce4eaa381ad965656436579,max,max,true,fixed_effort,,false,2026-07-23,fresh,101,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
anthropic/claude-opus-4-8,Claude Opus 4.8,Anthropic,closed,17,weirdml,WeirdML,coding,51aea1564ae15894b61f40d1011721701c329213be94dfcb63704b339a749d95,026673e30e070f8fa93b58716d749c1113c32c5318d46b32b6d0d225d4a5523d,016047521c58591a431809a45a440c6aeeadd415123b67fb28607bcac62c78a6,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,82.89,82.9,exact,canonical-source,profile_576644939af96eac30f5bf8d,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,epoch-eci,Epoch Capabilities Index,capability-index,af340960922ac09601323d31459eafa4be70589b149e667737626f22284edf28,d047418777dee974e2c25cabb67ba02b40443b1b29e9b7f8fbc57baaa02edcb2,60543c8e4496caff05af549fe069ab57d64806a3d2d840525d802b2d3ba5d2df,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,159.38,159.38,exact,canonical-source,profile_fc1802c65e18256a1d9ae18b,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,arena-text,Arena Text,preference,6c37268daf6f92a902ca7d67c0f189612946c2c75b830d83a1b993a9d1e31144,efbbd463b8eb0f704dcb1e1b41ffef3d0413121c17e3b05eea6e2c5286141bb7,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1481.6951264151023,1482,exact,canonical-source,profile_de975f255b440d53e37ca312,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,livebench,LiveBench,general,7fd6de2eff0817a6fb4f0ebf69039acb9535666a64c43310c5f09d18e7c90749,14b1744ee94213256cb5ee5bc5bff2f87eef9c606d78696cba601e3732472235,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,80.18829761904762,80.19,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,da050b7724b533a9c6096f7683fd23357b538b34e9c9028bfa60b7b55a8f4182,c4d00fa87778d3503e0da375a8106b8ec51aca1d5d26784025c8444e5f851cdf,5ff7843b5cc0b28f60ea81140d0fc83e0c7b0fb1123ff61449f0cf6367932ab9,release_5f3023448c2ae4632fe1a43b,reported-runs,public-multimodal-mixed-run-configs,accuracy,percent,true,43.56,43.56,other-supplemental,supplemental-source,profile_ad9dd60f09831d08eff03512,reported,unknown,,unknown,,false,2026-04-23,stale,192,2,B,2,B,1,A,evals-report-official,0,evals-report-official,evals-report-hle,evals.report · Humanity's Last Exam,https://evals.report/benchmarks/humanitys-last-exam?tab=scores
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,simpleqa-verified,SimpleQA Verified,factuality,7d099026f8b76e3d374df73a959e62cf580a0af7e6b53152c7e8f10b00ee2368,fdf0377c8ac0d9d9f760c418ff6e68fe6635fb46718db32a5f3beba5b2b11619,d23a2d7fa0743d0208611541cc6748e0c323c4e88079d09b9ec90ad86c184457,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,63.0,63.0,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T19:28:25.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,gpqa-diamond,GPQA Diamond,science,2422dc45bb8a8891c6df2dc782716af9046fcc726380512782e62ea8b11d8904,ebf0a4bb76d7ff5cd8c4f3fc8e90ebcb2c5f42803116e605270c16837708141e,0a5a76ac836c61bbf12c72c83694da49d5bac3d1cd626349cabb8bcf1a1956e2,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,93.182,93.18,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,mmlu-pro,MMLU-Pro,knowledge,c2d86f72d482f58a8d60de7323d37ffa8bc95cc4e566377581d07545dc836851,19c761715de3c79bc4d0f018115ed14bd32a079c58acc9fcfb0b46dddced4698,848266c1ca6001832b12dd64dabf16f5201cc770100b33e270810df234fa69dd,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,88.144,88.1,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,1855f1e78449892d90c30ce64afd6c27e69817a8f083409c65d61ad52e46c3f1,d90f993449dbd204972351c6ca14b2d1d997df4ecffb30569dc841a27d1cc075,f9d40076f0612e117e2b174313bd5a5f12bc67638455177ea0b9c021cc0b25ca,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,100.0,100.0,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-04-24T14:06:46.000Z,stale,190,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,frontiermath,FrontierMath T1–3,mathematics,692afc5427b97b84ca3a3706e6c025568f094d4b939c5573e8e115c9a8421292,a684bfd9dd78ed15936162cb91dab523970e070dc6227af7a336d3b82a61dd90,8cf0f33d75899c3ad5c42caaddf510069e92d52e9361c57eb73579d380f1126f,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,85.26315789473684,85.3,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-06-11T17:47:36.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,arc-agi-2,ARC-AGI-2,abstraction,00054b16f49484c2e7713f4ad9445be0bb6241fc3c0169cd9ca9e64ef2696038,067edc451b46b76f5953b715749839474d8cb2d750549244a1484a8fc09a1268,a0cfcd82b3d8405c11033eeb7ea6d86ff25e74d6a78b9934a39c06f28c60095d,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,85.0,85,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,frontiermath-tier-4,FrontierMath T4,mathematics,39aa59a0c81d86b6326b8153ecb1280a9d700266fb593688aaa0adb10750b461,5ed4709c602f1eab2d833fcfb8449250e151d2431986a420be1145514517d559,c6fe6d4b6009dd1d72f17d21a085b69ed2db4cd517d5f245cf16da2b7779c620,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,72.5,72.5,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-06-11T17:47:36.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,mmmu-pro,MMMU-Pro,multimodal,e33b664c7254aa78db5d782837358b6b06c4d232261575777ef0a87245635122,8932693ae8b5decf3b87081d435e09e1cab6b6de2d42b6b05b14deadd7d3ae8c,9211b7d5da0a6e6460b558fa72f27c3746d39f195003bca66df5041bd41af7e0,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,88.266,88.3,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,critpt,CritPt,science,dd6c03e533654c2eeeefa45748157c021d492ff7bc9f82f09e18f6e587465383,bbc78cbc55a6f9c9ab680114b91a6643cddc1d1f0d4beb22a4c2cb34cc992ea8,2896f47b263fbfd9d897702a037ae9dcf10431a99edf1fcf4e0b542f78ace5bb,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,27.142857142857103,27.1,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,simplebench,SimpleBench,reasoning,c5a437756643faac22270ddeff572f0880647c8ff8d562c9be341b85e9de4815,7809a4f77d0214dfe239ccb115abdce8f8acad87ccbb5306b8b2db63241b01b9,24ae4b1c43dd9aa74ec0e566041f1e4cd39e0181e5c0b9bf014e1d881e750941,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,69.0,69.0,exact,canonical-source,profile_cfd072d0392af1ffbd5bf34e,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,chess-puzzles,Chess Puzzles,games,343965f05ed05fd3e08277894113adf075a51bde4c90da0d878dd0d4c283e0ea,466379ffbdf4bbd523f767f8a86d452d561fde036b1b413cca763efef3ea5016,45b80790144bcf6294b80433581db10dbdb1c44a8fc359e009b98128f2eddd04,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,63.0,63.0,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-10-13T13:54:48.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,arc-agi-1,ARC-AGI-1,abstraction,64ccd242705195495c396c4d31d52e265faa781386828d62cb95fd05f7388ad2,5348b53e2e5d87b02e5a5126ed324cc210380c0e423b5dd6e50b5f9dbaaab255,b27c8f16f8fbb9f7481f43ff962202ec8ac65abac8edd3ddaa974eacb7d78b33,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,95.0,95.0,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,swe-bench-verified,SWE-bench Verified,coding-agent,360705a1d60be26d5d338d011c71eaa8df38f790ac464733880cfe075e66239f,3b1fb66d029a94b2c16a465f87bd85ab344db188ced7376c7d88b9e57fd09c85,d02112e899255831d29b3648680d2ac85b6c3715e3cdea5cfebabca486d41b89,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,82.6,82.6,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-04-23,stale,192,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,swe-rebench-v2,SWE-rebench v2,coding-agent,1229e757891d9bd44cf119f0014fe026cbe5a5f9f7263f5322e1db34b5a1f022,32b742003ccfcd025e4c3cff0917135cdef008a8f6af4242e8c943873503ae26,3d78907a601d1c50e69137936fb47907e786d140299fb0d62926c34f04f6ad6c,release_332cbf0e979daa9e173de9f4,rolling-v2,2026-05-01/2026-05-15,resolved_rate,percent,true,66.66666666666666,66.67,other-supplemental,supplemental-post-release-window,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-05-15T00:00:00Z,fresh,170,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,swe-bench-live,SWE-bench-Live · Lite,coding-agent,7ceef9410994e0d775461f97769558eb253249c7f40072af7f8afa4adcb30e18,20fcae2aa647ebc4241d94c61a6852bb9fd9fdfd76b230dbb1e2133224d70a0c,42e01a189064d8fb76dba75d73198590fe5d06830446eda18b634417e755a175,release_9d2de9d5be7e9822af19f7b3,SWE-bench-Live Lite,lite-300-python,resolved_rate,percent,true,62.0,62.0,exact,canonical-source,profile_f669e0173f71ff9f02fbaf9b,medium,medium,true,fixed_effort,,false,2026-09-03,fresh,59,1,A,1,A,0,A+,official-submission,3,official-submission,swe-bench-live,SWE-bench-Live · Lite,https://swe-bench-live.github.io/reports-0605.jsonl
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,deepswe,DeepSWE,coding-agent,32108dd002643a3c9c30777a8eb915c0a26e03f5f99d42aa1d69ff8d7d2ee499,68bdb8de629175dcfb2b193eacf397c5377ec172d2bd6730a1ae37f8c3e9fd6f,e069c3df0375db8dacb24d43ec010322abc67897fa4350149d14ef096dc0b9fb,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,67.03539823008849,67.0,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,gso,GSO,coding-agent,1c680b93356f552915b185d8ee39771fd3c5a2717a87648275ecf30d9c59fa42,0286d54442ea6771b6a1f6f81859eefb63f3f31562c3ec6818f3d615b21c18a7,415724002e0b1b17ef3d982a693cab73330cdb784e569fd3c433322353fe9600,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,40.2,40.2,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-04-27,stale,188,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,frontiercode,FrontierCode 1.1 · Main,coding-agent,6c3866e695389e6c979ca1944098d606ac5654a1ed5f093de33bce6e413369b1,99d132628daec23d21188fe1a79c9655cb69fab625ec1394e03fd3e309dab5db,a5eefc6dd41a3da732d191b44e47210b280e822c872fabb8a4c49fdbb88b3e20,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,42.96,43.0,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,livecodebench,LiveCodeBench,coding,73a3d39920ef6ff7722b4c18544cf7a904d2ebaabd344e221e0b0c0cdfca490d,a21764ca036a91fe9508746db8f3b58ba3a9abf5e60f0e2051c3043cf4ed7b58,3ef8da044c80b96e16dcae6982010b875e7f41534bcfce38f361590726c3ebf8,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,85.296,85.3,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,scicode,SciCode,coding,7c29a74fd429f966947da3f65f05fef3500d7092d7b0b14fcb5e2db8c46c01bc,28d3484f01ea7005d1d89773944b8d056e02f076f72d06982484aa781ee1ce5e,8588bbaea14ab90214d1b45223b1951a7c2fdeb5d5fc191a273e7ac539a17630,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,55.787037037037,55.8,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,ale-bench,ALE-Bench,coding,dbc13fd2569631e5eddad7bf9bfba8018cf20a5adc3d746a7d835507be8b5bc2,84377ef373b77b08f369b3476e99762191c9cad200a0e30a4621133eea97ab97,97b8488dc2ad8f8c25dd8467ee04388186fbce1e627f2e51de5c4985a0e11454,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1942.97,1942.97,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,vibe-code-bench,Vibe Code Bench,coding-agent,71f9fe5ad1a0c2ad7444178c5b992fb89ba4633f01736c73cb44d1302c362e69,ccbaf5ac24cba7f672e8107bc270a9138fffbdef5cb63c4116e0694e42a09284,68c749226bb0d2f4ea13cfe8ffe2ec61bce11047a1916876e0ed697ccf9c5edb,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,69.847,69.8,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-04-23,stale,192,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,webdev-arena,Text Arena · Coding,coding,1536af7d406532b7f6def9bd7397118ebf733788d378ef0027e8106fa10fd525,d2bc88e67f509ad51cd391016a75827a9074e7c1aa7ce3651c8017f7a866ff34,4b834aee64cad8078b0c63c8c2a6c1d915f39c020dc702e55a5d7df51fc343f3,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1512.099404756151,1512,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,mirrorcode,MirrorCode,coding-agent,f9408ebdbee4d9bc42392d14f2318c5b5d1435e7e92ab634aa5de9f1d0944da7,532e537e41d425835f299423de3f30f4b8ef4d1ff28d1e152366a3f47a38db91,c2bfd88c9a14d237e938d328530687bf8542ca96d58446303b33571e8d999098,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,10.0,10.0,exact,canonical-source,profile_de975f255b440d53e37ca312,high,high,true,fixed_effort,,false,2026-08-10T16:12:45.865Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,programbench,ProgramBench,coding-agent,231b34863da52d4808a7ff2a2d589c66f0a5a83d9e9eed712422301c0a0f4d5d,dfefc84ab8fd5fb08ed36c3189cec7a00d1cb2e580854c36b0df6b8e2743acb7,39c92ccb09d945d8aeec1e6ca91c474b55b831598548aa9a54147a1fb16b09da,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.5,0.5,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,4059bb7eac40e7cf74ed1f0857a568bfd1f030b34e46dad49cbf2acaf5f4c0ab,78cf8e0ee78432524ab0dd8ea14214bc3e6a4f2715fa58666aa9dcf9c0777c0b,af6c35129307abc2e39650ed53846263d385b5bd893a43db4d57f6a893781c48,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,76.404,76.4,exact,canonical-source,profile_de975f255b440d53e37ca312,high,high,true,fixed_effort,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,sre-bench,SRE Bench,terminal-agent,3e824acaa47d9d2df008fa06642c2cde6e183ce7bbbe51e45eb6587c4fe63b45,a7d5000e9a459427c0edb16ff8f0b7ebee22f0b97974574f38edd9ea0e41ad31,958b9df1a028a831ff40f3e6698200dd9c44511dbb9f3fe055f8302bd3a24de1,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,3.817,3.8,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,mcp-atlas,MCP-Atlas · Scale,tool-use,0d2722a0395a01843464c43792bb7608cc38b927eec0f01612572e0c3c999d99,ef573426c45bc2c17b6e85f47695420c5b81e29bfadaf7501ea1fe3ac06d68b1,2a0988bce9142ef1003f14344b34bb993e5589051a2304bc6c7c78694cd2c3cf,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,75.3,75.3,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-04-28T17:25:29Z,stale,186,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,skillsbench,SkillsBench v1.1,tool-use,2707b6e576b5cac0a3c6580eebf526a0cd6c1e21b70b1cd7f492d2cae500eec3,481eb250f485c83e56664ebebd379a05f19f8fff08010639e58cd7b4680ab523,c2ed2025e534826e5e627446f4493a128269710491b35eb2122abbaf19d56304,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,67.3,67.3,exact,canonical-source,profile_815be32a4499e3fe8b5cd1a6,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,apex-agents,APEX-Agents,tool-use,1d90a4f78b0883f8ca734606748753df5ef7308c1d41ecb25a4688ca7d6e4635,3059b55ca9032be8fe5527548b7f86644db0864f87a2b12ae144f485bc44ff18,8f571d2f323b02e5358050aff91cbe030cc6ad0e6d26fdcdb64f120bf9ee8286,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,55.1,55.1,exact,canonical-source,profile_aa4c4f569e5bcfdacbdd9f30,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,vending-bench-2,Vending-Bench 2,multi-turn-agent,963fda16709b38b5209238cfa544ae8b95af884b12f7196f30efb69dc1628f46,a1f25a63b5f90ee3582771319dbe06cdebe5fe77f3c2cd799ab6144727ff8c1c,16dcef846831a895f1270680fca37a67cd20d43510a0e7fceeb421430b42eb58,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,7523.843999999999,7523.84,exact,canonical-source,profile_e2ab7970e7428fde1eca209f,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,fba02100adf0dc88df150bf963b5f736a3d39fc6bc32ea15466c814f459a2d0d,65f65ee31ea8aa2145255fb8787277aa910fed3c1c18cdbeaaa0eb3601e57219,3299e6f51d4d3a157382e1afab88327c84b1da7d239dc86c20e50a38d82fc291,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1359.24,1359,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,osworld-2,OSWorld 2.0,computer-use-agent,12f26363df699d6dcd603249a996579b3b96146e7ded563d2252f58b61d24ba2,8ad273c5f514f7ce2d652632e6d6a0cd8c003142afc8f277c56908a51d76a495,a35a7e4f1f52502187e775bbdee0cee04878e831ca9612df4ba9c8ae53cebafb,release_15c5525fe39f33bc181dfb69,OSWorld 2.0 current leaderboard,binary-accuracy-500-step-budget,binary_accuracy,percent,true,13.0,13.0,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:27:06Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-osworld-2,Epoch AI mirror · OSWorld 2.0,https://epoch.ai/benchmarks/osworld-2
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,fe7cd6d9a44f442e9316ca81652ce89d1c6007e7ba7fe61f3723d7dc87c30e1d,6ae289ee0c6ee62514bcac78f01b51628dc7e7c8f0740de4e91366fe242e530d,ca1b0f52e5f57f2a197d9554fdac27ff8518c746bdb61a2a89fe030c814a6421,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,44.58762886597938,44.6,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-07-22,fresh,102,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
openai/gpt-5-5,GPT 5.5,OpenAI,closed,18,weirdml,WeirdML,coding,b0302da98d4f6fed9067b3c73c059f810e0e9cec172322ef11c4e0c437bf83aa,388d858d86dd99a3cf0aeb3bb4c4cec793fbb5d92a94c42e309df1d38a633b5a,5f1e963ac2dd299059fdcd04c621362c1517cd74eebd1c8b30c33821cdea8047,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,84.91,84.9,exact,canonical-source,profile_618796ded36fbc2ed30584a4,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,epoch-eci,Epoch Capabilities Index,capability-index,37fa9fa7b8e9e2663312327512b201f9267a50f57eba5db712a5dc7c8ac37f2f,93143eba8da68f0639cad1ef6081685f9b895be87f6f6ce8121e45980be72d27,adc56dc177f1e00a3e47615980d8b809bef7e65a79026a6f4318576b63332afb,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,158.83,158.83,exact,canonical-source,profile_2db9ce730bc98e27173d249a,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,33df0a073fa8c21c28172b556130da9f8b8e7225480584ec920e421a7854a7b7,9a7349a79738b1a52f9fc2b8c5c3dcc7da8059bb5d1852454f24a731fb71cbbb,7c82f2ce14313e9c37348f5f517fa708a15246e06672c9ed62859727e46f7218,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,32.489639,32.5,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,ccf4cca8359654bf6637ccec3c4045a9addd0325153de7615f5fa7dde12f6ebf,d5d55febfe8eef35c95a238b2b5ed03faff142bfefd2207322634cc0caef0650,be9d43810413d30c86cd140f91203d82325c7eb5bdb9362db23ce77a0a066295,release_fa9e908ac04b067eb00182d4,HLE official leaderboard,published-leaderboard,accuracy,percent,true,44.0,44,other-supplemental,supplemental-source,profile_73c7903143f8bc46fe1fb626,reported,unknown,,unknown,,false,2026-10-17,fresh,15,1,A,1,A,5,C,self-reported,0,self-reported,hle,Humanity's Last Exam,https://huggingface.co/api/datasets/cais/hle/leaderboard?limit=200
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,7198729b47a30b6b8f78bbd4ec5facc7be510250c207c1803d972d73c1556a2e,a4d1e5fc51971db5cf57c8b18bd698f7874498e6a43313b64e39cbf37bea0e21,33c26e1d10e09ca503f161d3c19688771af0d60c45d49433b41263e57c564428,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,43.327155,43.3,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-19T04:43:19Z,fresh,12,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,simpleqa-verified,SimpleQA Verified,factuality,4898fa4b78ac9134049cc74fc7dc59123e5f3532748749af8ac3878f5c1887f4,db15136d32a68252621a859d398642728481b97a9416cee0822da724f7268e83,99cbe422c840bfab4fef49ebbebf773febe7cb793e6b2ce2706556e05b6df312,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,42.5,42.5,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-19T22:38:54.000Z,fresh,12,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,gpqa-diamond,GPQA Diamond,science,593a054ab7859f4e0e41b82917e41f6145db7aca7aba0bdac914c7598994f5ba,0d235601c73d45a8648243239a80ca13217c89c2e8c1fb11c706e28589b9c608,a993234f8f2a31d3e51cc72ccbe3fb1f1796cd92c777cb68a0611427f40baad5,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,93.18181818181817,93.2,other-supplemental,supplemental-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-19T22:38:32.000Z,fresh,12,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,3798ec2265dfa45bbb865b43d9adf1dd28004f66419405b76cb9129c8b3ad22f,1b0aa3e2ca7009bc8fc138eeb6d792fd40a9d0dd129e6459e96ccea690a7857f,912facf3dfdd839d89f365cca3299dfb731a2abf7798c5e671fbb0c2e9c8715c,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,99.44444444444444,99.4,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-19T22:39:00.000Z,fresh,12,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,frontiermath,FrontierMath T1–3,mathematics,4416743ecbd34b5942afafa8319d387f36cd4f179e4381d5c2dabccd9278576f,0ab38c20c2c4d4af6b3a20b4d00c661d7a45bac153ec382e1eae6d3acbdb7e61,3763a53937183b88c672c91b1415e69144f5ab19162000326b0d89e7e4592b97,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,71.2280701754386,71.2,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-19T22:39:00.000Z,fresh,12,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,arc-agi-2,ARC-AGI-2,abstraction,26bc373603540318270cf38332dcb9bca941f44f00193ec2716b827f0a359b27,81c77fbadacd876a75e2fd78260db02b745e81d8f99c0381655d77d6e3a63dfb,53e1415093a166d4dd0cfaa6bba0e34840feb18b462366fedced90269d47296b,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,78.05555555555556,78,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-21T22:23:16Z,fresh,10,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,mmmu-pro,MMMU-Pro,multimodal,cc68d471c6ef075db820835c8bfd61169da3b839603e36b8c53726296ec251f0,6f394f9d1ed54c6373d7dc8bf9d211b8d591247cd60be3b5d7460a7f239ed3e2,1eb33289ec2b6b80c3daa249d86058c34f4505c520be1868adbde17c2a4b3970,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,80.635838,80.6,other-supplemental,supplemental-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-19T04:43:19Z,fresh,12,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,critpt,CritPt,science,f06928c1316e7b0e692d4bca43c971fb48e25ea618f268729e6c74b72a74ef1d,da672aa940290063ef23b77fd3083ee2c01b1badfd425062a644b700c0c49aac,e5700bbfa0fccc839832c4b41bc0dface1658207d6ff0a083bde7c68aafb9457,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,24.0,24.0,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-19T04:43:19Z,fresh,12,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,chess-puzzles,Chess Puzzles,games,7b5e008e284494132d8eada58016cfd9dc0bc46c1743fd09c0eda9300784f23b,13595905a376c6a3112853335317d9134bf6ed5bd28337dba019e67de9d70847,52e6068ddf31abf8531b046e4f0709763137eadfe121bfac9c15bd189ad1c2e5,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,46.0,46.0,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-19T22:39:00.000Z,fresh,12,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,arc-agi-1,ARC-AGI-1,abstraction,991c207445e2a747e417f7248386683fd76f95b830e427568e9bcf9e053a937c,b5a191fbde6799f7f5d14b6f700ffcd972788e25aaeca72696da3d96f4c12066,c390e152fc5ba816c546d9361d0cc18e63f1f376203c9c100ac914b3606c2f23,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,93.5,93.5,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-28T21:25:03Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,cursorbench,CursorBench,coding-agent,63945ed15e41f2c6971613acb391e683fe2640728dfd795f1ddf2ee220be7e81,b5923d48eab48194bddebef7f147c068fe856a23837c7236eefb6db26777e572,6c641bc1d76715a3aefa356f3f8396b2fa8451327c856d7292bf3c0b285d0282,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,47.7,47.7,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-28T21:25:01Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,scicode,SciCode,coding,bb110ed026e0f841fe9ce59709d6a16c7ad9be26417db09892bf43353d99c849,3611dcd3d406a56e3a6c39f01259b47b3d0c99285138e75d3c1a5e226000421f,60794fc5d94f43aaf01512adf1f24b963edafcc67c7e7b78ccf9c642648ab819,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,55.9027777777778,55.9,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-28T21:25:02Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,ale-bench,ALE-Bench,coding,a8ff0a061a863b3c0d1e951729de4bb8d389c35f2d4f355c59ddde2680b76c1a,08478edd08333a32cd0555b8c862a6141cd24439985f9ef70fbd0a9600f2315f,08fef3acbf0eff28af2e87a090728c27fc19eee807ce11bcf5068011d98d89d0,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1622.0,1622.00,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-28T21:25:04Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,vibe-code-bench,Vibe Code Bench,coding-agent,6e94a7ac8791961c0bc57999c73e7442f35b47232d862308bef9ae6136b3a786,ef8ae8eedf2e8149f10fccfc7f13b4ccfe3516539c4805c6a2b9c91b2805373c,416d0e02836e95692eca6edabf7d15cf56a0c3dcfe669f7148325fa3b55207c3,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,87.73,87.7,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-19T10:28:59Z,fresh,12,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,webdev-arena,Text Arena · Coding,coding,df43ab20c370c71592258d11b1a61b79067f7ad39ae7d15202f251995254e397,7a2ad2c89e1486db79f897c1a76b1795dbbb191905f77554071a73109ec2e1e5,716d55c6e96665a5c9562d2f1c0d93b6704a47af5d653fd6484634137d6cc455,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1665.040445245608,1665,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,weirdml-v3,WeirdML v3,coding-agent,1eb90cf001e40407efd51de25bc6078c7bf3970961a6030b0afe8a90cb158570,e56aaded4fe4dcb41d70ed5a329768946c3d1990c5dcfb7e0ee11c6f1d33b582,f270043cea84f572be6ff64750bba9120d03125b0f6fd0cf2e007fa38472df6c,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,9.19092568181818,9.2,exact,canonical-source,profile_c944b51d250e47008d0a1b60,high,high,true,fixed_effort,,false,2026-10-28T21:25:05Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,bd8874b4f3cfb7a1b6e5d86fd7e0bcfd54b034bacd8f72e45c68df12665f08c3,49add2a57803c137457fb01c18ff9f38f225c650649e99749cff2e43d1511930,201ea99257996aae4b74b7d921b921ea2c21c5a41a6806c971b39eff36862282,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,31.818,31.8,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,apex-agents,APEX-Agents,tool-use,bb63f4c2807e64f3e0d66c16b9bc87caea39ba2480b8f94931ccad80de517bbe,afc9ee6bbdef2c7a7e8914d48ac8fd08ecaeb581fc5e82c69b7597c359fa9a49,e33425a8ffc0b19b0e6e238eaf01bf5d3925263dab4b2ea2d87565b486b29226,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,69.8,69.8,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-28T21:25:02Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,arena-agent,LMArena Agent Arena,multi-turn-agent,a57d2e9f7f346277c78493332710b5f26dc59a5616eafab96440302bad2b0202,295aa6ec06617d78dc1f8d0ddf7a7ddb35ab71319a38112e17f3f694e8a11f48,ab294d01dec1bda74524da3e7bca7c06a30db1445d6d21a2f63bc8eaa189a260,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.026748579776335085,0.0267,exact,source-specific,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
deepseek/deepseek-v4-1-flash-20261017,DeepSeek V4.1 Flash (2026-10-17),DeepSeek,unknown,19,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,70d58da0adaddb4e905eedd75a199d295cc4e3a1b6db2c273bbc72d0c290dad7,ec55fafe68c79bb9a404bbfb4c15a08e08abd68cdad13eec551e200dfbc5ff66,38ca9c850a82045ca3916b3f20dab96a791a9ccf17f6190197094bf8be599d7b,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1745.77,1746,exact,canonical-source,profile_e01929ea31f86853f12a3cec,max,max,true,fixed_effort,,false,2026-10-19T04:43:19Z,fresh,12,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,epoch-eci,Epoch Capabilities Index,capability-index,61608c3f2a0f959809c282ea32b0ffb301246cca584089e356bf2de16f3aa5fe,348beaea3157ca3a7c9554b4fa06744a1f4047c19ce6a78e8a7e4bcdff389d65,229801ef2b9d4fb28dd8cd63272c855c333d194626843b78494c92e218213ba1,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,158.33,158.33,exact,canonical-source,profile_a2da5036f8e35f4973a1f1d7,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,arena-text,Arena Text,preference,480c0d43a3d6f866625dc0362464772612568fe0e29e3764ef134302849ad711,9476ac8858119b35b2b49171ec79ddf957d037e836af955536b1dcb15c1e09dc,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1488.1894728804475,1488,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,livebench,LiveBench,general,ac4d3bb6ce98793523f1752d6409dd79d90903fa09298256eb2cb10602a05638,0ae44147578e5b57b250f7990a149b0ef43f36765bdf7f3a0537128525a5febe,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,79.19304761904762,79.19,exact,canonical-source,profile_2ad36c6efaf22e10e2082fe7,reported,unknown,,unknown,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,f72fb32ca3e8153ca58cec8fc8854dca8ab7e989fc9a620b6859ff79f785c84f,e35a8b6faf99dbf6a2688cb01c029a24c7760cbb408df0c9348eeef2bd86d431,6547241796b7257e8ab884f956a879819be7729074e34228a3dfbd69d4f79e1c,release_d43dc12b725969a1cb1e2a28,HLE-Full (provider label; revision not restated),provider-report-no-tools,accuracy,percent,true,43.5,43.5,other-supplemental,supplemental-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-07-27,fresh,97,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,moonshot-kimi-k3-card,Moonshot AI · Kimi K3 Model Card,https://huggingface.co/moonshotai/Kimi-K3/blob/main/README.md
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,fbf2cd0e85974db77b1116ff09341f15a938cc00a496710870bac8a7050f9b7f,35788b0c6ee53a32d8a5496c7eeb81c94ad432d49d34877ed89dbedd2574aa99,e4222e9089f31e296bed718769c7c24923b2b7499edffbb3a5bced24d21c697a,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,46.895273,46.9,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,simpleqa-verified,SimpleQA Verified,factuality,c31873775273f98c9bee5cd8ec3c78648bcdf8ff588874e3b4f0ab58cc2691e7,1d4114e7a5cbf8f28dad6d78ae990e030fa9cd9ab58c24445fdb06105aed6a0b,8a3839bf844fafb00603a5223443e491cfff9eb6b96d800ff480fcd85634633a,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,50.6,50.6,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-08-27T19:36:12.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,gpqa-diamond,GPQA Diamond,science,f2ede0b7f010fe42dacef5648c07fba31a85b4f5234b3e58f800424b5c5e107d,a792b2bdc4d8fabc21c787eb54aa5697730385a38587107f4ee3d75b9fbb6474,7f3c372b5b9d08c79117a9d056e7d7c5e6f0a18fa77ce40b4966a6d6caf74611,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,92.929,92.93,exact,canonical-source,profile_1e5c3455b6099d35084a3c99,reported,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,mmlu-pro,MMLU-Pro,knowledge,59426b4594744d03b9e04a03a5cb588b729bd0063603f8fee7b15cdebf52147f,12d3d378a65be702d209dc0713f9eccd92c10bd7e32f02b0e374efef379565a8,3039f87c799cc0ee74bc38b42ab8172d0e1a0c5d9afbcbbe899e385734167535,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,87.973,88.0,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,6a1c483792d6cf4e01878822d98d05c08322d452486a23a678b82ff2bc61da3e,16b04537c155a1bed20eeceaf66f7bfba282c40ee2b0c90b2d084a2c0c96ced6,5083290679b6574e185789e24ba694af1cd7e60a77fa6d03706ab192a8c43cb9,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,97.22222222222221,97.2,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-07-16T21:59:40.000Z,fresh,107,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,frontiermath,FrontierMath T1–3,mathematics,61249d20255670984ebd7abc1f9c7a49afd725bf8657e1d3d528a57c5903540b,7c1a329cf3e0ce523357e1a85d405b34828175da2d6bbf04c7daf5bd7fa3f6af,8cd6dd3d3cb2d972401111c3a3dbca3df50385d3b1f06d522a56977e1ea3578c,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,72.1830985915493,72.2,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-07-17T01:48:42.000Z,fresh,107,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,arc-agi-2,ARC-AGI-2,abstraction,4a07c96ca40cb010ff958abc0a217311717f944c74aafe903973e8ae50cd9981,9a15dd5e79c0a13ccede080eb3e9fa564b924e3182cf6dbf90360a34587738a3,42b22b8742fff641dc193a46a202bc267ec6e13ac30cbe53bf8a3e841fe66b6d,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,60.416666666666664,60,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,frontiermath-tier-4,FrontierMath T4,mathematics,9931c4d00616afca783048d1d265bcb1eb6f7f062be8998a37da5f0f1f6560a6,25d1789291243ea1e2fce19c790995733757a28597d19634bdbab8d259817e0e,6a845d022df3285937b93ad8eb42b1c9fe0becf0699e3a92964e5c4b4d9c122d,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,39.02439024390244,39.0,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-07-17T01:48:42.000Z,fresh,107,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,mmmu-pro,MMMU-Pro,multimodal,0cb4c511987dca9997bc8f024a582bf9584c4fc84b60b55c103eae2ff283c910,79b7dd03fe7446350d7cf01939f3975bc1b976dbd507b05c6936bb0e3a04d904,6919a5c7733cd4daa2083f67ab0e7562b65969b8d1f621fc3b5e57be2792b6e1,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,88.15,88.2,exact,canonical-source,profile_f82da36312f78d41c5384ff8,reported,unknown,,unknown,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,critpt,CritPt,science,b023e07b057566dee92d055d89f73cc9174f9a9b31f6247ef92983175b721b82,4645273c8a6107ddee46707eb8fe6602f7f1af744df83d796ee9c6d08cbada8b,59cca39d88f18559fc0d18a44f486256be02cce06212032473cfc39f9b4d867f,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,23.400000000000002,23.4,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,simplebench,SimpleBench,reasoning,eff8b1e9d50a3b1c08a34f9c8ce395d35fd697e9d6adfb26c44d8e242b07ea83,abad372fa85eae820b735df13516eee59337c51be75a2aa64a3769deb144986c,34d15d8fb21b57acc8ecf80c7513e58a21cd3eea609126bf88cf0714279cba0e,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,60.70000000000001,60.7,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,chess-puzzles,Chess Puzzles,games,ecfdf4fcdf6d3a488d63e466e9a2969c23f224fc53c2a19f4b02c60340c01a47,ff23b72e4a56c00887d77c38655ccf87e4f6b9165e8196e7d915e43cade47ffe,8964ec3f7cd10291681183923c3a7e640253d9e24a5db073cd0846202888cf6a,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,36.0,36.0,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-13T17:13:35.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,arc-agi-1,ARC-AGI-1,abstraction,ed0e7ec4fb563e9a5f4dbfd972b7963a4c47e6fa6880d05748b0d772c05e0afa,fd639615119049ba425cffed797db19ea3a277a0a075ffa5346aa6c77a90a942,c0f8b1c6d5c3d190821c86efc20cca9fcd013ca7bf67de5beaa04c177ae4459a,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,94.5,94.5,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,swe-bench-verified,SWE-bench Verified,coding-agent,cfac77f030aa5416b30a79c99594184406dbc65e80d8169f09134cb73366259c,84004b8d1e4e0c557dff82cd3f35025c4a609e779c5ba85a2dc3b16e13bb25b8,e5146a7e1aeb95ce4cf2f69911cba1d5e88fe4dc3a17a5efbbaca38e9ff01ef3,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,93.4,93.4,exact,canonical-source,profile_cb71961e616ff6ab8428806f,reported,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,deepswe,DeepSWE,coding-agent,c25ce5401f264cfdc37f52a96502befc2ac014b482503be496e2a1d189bed513,7e47747a62e403f96ee540c24b9dcb6238b490abd78f21d7a73484b9d6d730f9,306487a785d65dc89a1708ac4b32950b6ec9585e38149e470860ea5d905e336b,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,68.51441241685144,68.5,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,79199cd5aad7787304cdf9c1be8893b2008a676b2aef5b72a6c46eb9d4d82dc2,5a878bc490d617197f52a1e75f00c2ee7953c7df20810df4f0bb3bce39bc3174,e37daeb8609b20a3609fd19288963bbd30209e2a10482cf29a2dc7b37be58c7c,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,97.7,97.7,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-07T01:44:58Z,fresh,25,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,frontiercode,FrontierCode 1.1 · Main,coding-agent,a03f0e262524f20dc1f163683bef67b8ecbd4202dbe3df9aa52658cc8942beed,6be974353ee3e7502cd03aff9b94bb8322b624cdb0fb733d576a655fadc7b29a,c7e0ed9a1238dc923792161e4c4e2f6758900ca23d2261dfff6b838e57b05782,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,44.17,44.2,exact,canonical-source,profile_1ae7fb380652700fc7c15f73,off,off,false,none,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,ioi,IOI,coding,c5c941f2ef3fdfc2acb4243f320b21b2d11491b5fa8048370fc02d6653065d5c,1d9bd114d0b268b05f2f31f6d8f3a5b26a8b9cbbcc7b39f89419ea6b4e3fafe9,d00e0d2333d717ddddca461635327b78e3c7fd75b1ac49a6adea50bb44c02e28,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,48.944,48.9,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,livecodebench,LiveCodeBench,coding,96aa2a5bd779ad3d833ca401a08c4dd44420e88bfeeee61fc72d14a95a891831,a7ab42c47f4d373b685eb27cc1ea7350b0b957c0bfecbe87bb0066b301eac89f,cf909926925b282a11e854233cfb493e4c415d491b923b787e755d2b23db3c7e,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,87.185,87.2,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,scicode,SciCode,coding,02cb5ede8aab72153b450e350942ca8b5f2895ec40ca1e412ece808da6c4c8f4,a4de8b9ca28d41344c2995eb7d0554c72c2041b8afe7d2255edbb35afae068d5,d207f4bd9a8315a188c36c2f8f48954a11e30145087d14971eb8a55d67f6fda4,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,59.4907407407407,59.5,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,ale-bench,ALE-Bench,coding,0eb2960b5b30da0327f912a119eae92205d20e740d345580ad55b44e98e5a9da,edf2b44f7924b4f0346465e2a3b59f4c048ae89bd1f3b8ffaac491b647023387,26f8e2ed3572372625e53bc30a10bd0c1ddf914250ae5e5186f642f3fcd07773,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1524.47,1524.47,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,vibe-code-bench,Vibe Code Bench,coding-agent,8f2f0b40b1ec183ede22e557e03ab89739d83f310abc1500b494576e965bc8b8,9183c9040eefd31a71e9afe2a7928cd78a69fd52448d64fa5e9d89bec7371ec1,f23420397d0c64bc27c2f7696c7528bb4ac34ca86854b63260214860266017fe,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,84.968,85.0,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,webdev-arena,Text Arena · Coding,coding,fe3163e47f8e0b86b0faf8abc088c1adc6b1667fde0f3ee046810a3061eaada3,28af1c77bf0f898bcf261739f41391697e2bbd32bb7e7de9fc935cb1abbf0b9b,501d10378f60694547bfcef5359a990c3ad47ec25d86d2292e96402e7f52b280,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1656.390908396489,1656,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,mirrorcode,MirrorCode,coding-agent,398ed414a629db5e8e68d52f8899167266f5c95a3465ed8debccefcf1c93a662,d13cd7b119ccd2939ef1efcc9f7575277b4ba66243b0be7a6d0f39786e3daa2b,904c442ece3e21ed6442928487a8e3b1acb439ceced20523cbd0f0cfad02f58d,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,54.44444444444444,54.4,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-09-03T21:02:17.000Z,fresh,58,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,programbench,ProgramBench,coding-agent,26c4f5a61034be49e0812cdbdaf1ed3b91bb046e924ffddad65fd58dc09a3e7b,1199c5fc0425e063437979897efc4f898fe2f020025744309249e7967fa193a5,f5cc6678464d269804314fae3c938893505b15b61747dde2248eeb3544341176,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,2.0,2.0,exact,canonical-source,profile_e69c6787d50e591da58165c2,reported,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,weirdml-v3,WeirdML v3,coding-agent,9a3e4122b31e984afea2977fce64ec66b801aa2261952c4a5d7857ad10ee1537,9fcdc050713c6704d25200133c217bb92b9bebb0870e957041711aa17dd84237,ad05b76231e6475395cc18578eccee22561c4b69022d7d5fa111eef6a70a9dad,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,7.544402727272727,7.5,exact,canonical-source,profile_1940792474091774d582a01e,high,high,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,c7a9ff6cd8ac896fb0c564f09d74cb8061195fe5ba83686fe507a418bfbd05f6,f011615d3a0ff4fc40fab8b7d0f16b30804a5802595601b1c721af7056e577c4,bddcb574e9187cc511751527acff9b2b854391d966cc657335f1c10a25bdc9b0,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,80.899,80.9,exact,canonical-source,profile_3a3b2af30725252e205c7697,reported,unknown,,unknown,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,a1958c0ec515b7d37bc1165633c9986298377fc6a0f85d68bf7de7a2cd1ee0be,751dcbb025c180421b819945355a445872c5d6bdbd6c4eabcee4897b72fa9d64,55c2e28ed2702b6791c427422cba51c5f0c19960cad272616f5e5e584f850d68,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,17.172,17.2,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,7121334ca637d63b80fb4947ea06884ffd8125c770abf8a12567dbe02039780e,dd6944447fcf58ed7af6a011c94b94c9d72cb4838ed37362ad0c9764abd4c293,5bebbba10f9c82136b1f1403f12846611f297a8d15029ddbd52cae0338fed613,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,45.979381,46.0,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,mcp-atlas,MCP-Atlas · Scale,tool-use,207375a2fcd92f0a3962173fd1320057d3dbc61525e200e14dcfd6924d63f3d6,5bd39f626b0e2c36f82d1e635c4e08282499652022bef684a03e7c0625636b16,8f8ef31c71c5ce67f96072b88c117227444f74d83e913e784c3cb33d28051547,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,82.3,82.3,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-07-20T17:00:18Z,fresh,103,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,apex-agents,APEX-Agents,tool-use,b4feedd80a1c9f968b0c94b651d26b8bac3e7ef3cb7bd56b0b682cdd7e3b1335,a777c33985ce5ef2262b7d7600f1e8b80638bdeaed19384aeebc72919e68b03d,70143db5f478d32176a20dc8404621bf9500598e0c746d321db264c6ab9f9ca5,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,47.8,47.8,exact,canonical-source,profile_9b48cd9b9e58065d050a20a9,reported,unknown,,unknown,,false,2026-10-19T22:24:14Z,fresh,12,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,arena-agent,LMArena Agent Arena,multi-turn-agent,585da9e2b1977e98893a98471ae6fe1522da8314260d7e10398f852409f37e66,242b341efbd4738c3af01d4048225eb3356d717cfe6178606f786c2addb1565f,206f50383b99286af4fd22c5271f21a330a236359c338c3cd1968c63c6ebb157,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.010787136728542578,0.0108,exact,source-specific,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,vending-bench-2,Vending-Bench 2,multi-turn-agent,93cd43098cbcee209cbf88c02aa3099df718e71666ec7e80b974861a4a3bd8eb,8e292c0ff260eb003dc47a7c1464481622b8aebd0f40b2ce93c970d629bff307,4773551f57125bed55432f6478c79beaea73730ddb3a7beda378f762751e4d9d,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,5165.035000000001,5165.04,exact,canonical-source,profile_97cb45dbe3d881ac6db37643,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,2c4b28a007cce0a843736a7cef1ce174a17e42ca820a43ba099e2af985857475,0bc1cddbbdc14f64197687f8444144b9383302dbfbfb78a4d46018a80cdce725,59aaf3d6f2f9d29609b06bc1b843ae84fe3f8071c91b96c5a63bfb3106816c3e,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1542.76,1543,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,2b9661fe6fbf9e70046297d8c267400aeee5247e9dd5e15d499a65154e38e370,5c1c738a03a0e413db4f979111ccedcc13da3dd529fe7ebde35c5d53d215620b,9468dbba5cce4b78e5e35f6b5fa0e8cfd19ed4fabcdb56a4f4f8f6e3986a8010,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,1.429,1.4,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,e1f9e1614fd456480496d1488defafd0e128454bc32dd5116ed296a41447c0a0,e81a9927a3fd0b940192acd79b954187e22d5b7e3753542c3f53f54ecad7db9e,d4add4b3535a169ff2af2158c9e2174929b38e4bf3719273a4645375b4d4410c,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,37.11340206185567,37.1,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-07-24,fresh,100,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
moonshot/kimi-k3,Kimi K3,Moonshot AI,open,20,weirdml,WeirdML,coding,c5c832f089347ca98accb8140afe4eb69f3f72a8aeed21a767dc002964c53a4f,52e0014ebc7b9f4bed34ae9119c7ca48183604808d9157ab6cda1fe9aa817e52,8def5ce4f9c29d1ceb1fbe271d3ab21fe88d2533a452a1ddabbfb9d9427de49d,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,82.57,82.6,exact,canonical-source,profile_2de1840141225fae3ff55b20,max,max,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,epoch-eci,Epoch Capabilities Index,capability-index,896b94b91a3b7fe0621f037d89853a5e7e3e030d7e0b188930c4518c46cf0546,ff31b29d3f1268d878618663bc8f5ed9257ed209b2e2f3b92efa61f043bb1e1e,a4c6e5b59eb818d5e7a4ba564b3a76f08f622d4a6dcfbfd37890bd3275ea41ea,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,158.17,158.17,exact,canonical-source,profile_2c8a392b509c1c0521fc9bf3,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,arena-text,Arena Text,preference,dc80e57791509367752390733b28216679429620446dd0d2a874e825d3c3c85c,effd0ae6a372eb59d5e83f5c75f693daa82c4b4728d9d3362eab1c8249095398,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1489.5500575529784,1490,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,livebench,LiveBench,general,0ee8ec45dfd3ab0bb53f5df558b16a35bf9330b031f29cc9efb69a4f0d90936e,61e19f8cbaee775955df454a93d83378383cc1c139a43d4a6b56255a7c7460f8,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,78.8260119047619,78.83,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,d98f94c11efcf8987c40fba35164fa7135709ef3395426b13257694f6e62a47e,1c45f03a59a55567820a001236da0986e05bd202fb01bcadbcdcc34dd8527c6b,ca7773258dfcd0ce156963016827db6b1137cdd4c4f0777571ce92f9798bf5f5,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,47.868397,47.9,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,simpleqa-verified,SimpleQA Verified,factuality,1516467a06c6c43709cdcdee0b4f7d7b3356c7a1c7311ae143ac10d116539b30,e80e3e5cefe0907999bbbe5520afd5f232e7da3bafca2c799bb7e3e275b560c8,9f5799cd60e18ca9aa0eecd6743a82f7e7c58c7b690fd09e185585ab33e03352,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,69.19999999999999,69.2,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-08-27T19:30:25.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,gpqa-diamond,GPQA Diamond,science,f9793a6432e4eaf9a8dbbc173cf047dda3e6ea783f16cf2b55d6dc940799e7ff,125b03eac5e605cf220ce84dfaba9ee5a5b8c8c64af429206b1bef66acc40a10,29c8d240d75b1279a0d31a1eb0c660c187be5bcb9848d9610e321635fd176b35,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,93.94,93.94,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,mmlu-pro,MMLU-Pro,knowledge,3bec72c01588149c3ac00d94198277f69beed60615554951b770f7eed2e4445a,09786e4445c4b228e536ab725f1d379c1471f0702981dd01d0c6468255e08538,60fd7fe0a42e91f4b9844276aebedc209a9231e29d526fbdced17103bef37018,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,90.121,90.1,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,5c350c11ecd01400f7406ad648e69c0683fad881fedd094fe46c8d7c5c417b95,c98de7d2ce5590230fa9f7bf8153fd3729541ca9e000593d4c6afe7d695d58ce,32f8df7bfa32c631c95c02add8808505ef388aa6348f5af1da1f0fd50cb3c08f,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,97.22222222222221,97.2,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-08-14T18:29:12.000Z,fresh,78,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,frontiermath,FrontierMath T1–3,mathematics,dd2d1fb30d9ac3c51b1937bd1e104be5a57f9a2816d7cae0403f327add08f579,750f6de0176f1b01bd9c79ed1c464a967f45ff29c0390b58e5d64e1a6da7d48e,5c4ffbaa4694435d16724924bd3e92e829ab6da23f61d6ca87985f02225a6a87,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,71.57894736842105,71.6,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-08-14T19:59:58.000Z,fresh,78,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,arc-agi-2,ARC-AGI-2,abstraction,5992b543a60fba78c3638cbdb97f20afd78d647286b01f8f6dea8c5707568bf7,246e7bb3ace5e2ccb7b30bb4831ae4a6d02720d2b447125b0a8d622df1a99663,f6c9bfb4f2fb48d9291daf5ef9fc0e68135afbc6971eff8c25aca2f196043cbe,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,84.58333333333333,85,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,frontiermath-tier-4,FrontierMath T4,mathematics,d8d95018d3032e0927e0c8cebf8c70101081e3318a4c62d1867941addee77562,aece6c458c18c1fe2961dffe8d146f0274ba162ae7066d9e8522335ab79523b8,4c066fcbc43cca1fab40e9bcc635e96bfab7551d349b999e087229dcbb8a284a,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,36.58536585365854,36.6,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-08-14T19:59:58.000Z,fresh,78,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,mmmu-pro,MMMU-Pro,multimodal,3d1a5f00604391ff36b286235eea165e4a6967c2690057884f284d98e2f87aa8,8844c58ab235fe1a5de98394b8a16ea3414f6e8522a1ebc662658fcef294d9c9,5417bc3335af1f333eed959cabff2962ea36d6080781b040aee9c79a6bd35db0,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,88.96,89.0,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,critpt,CritPt,science,96810ef15bf82fba37137727f3077daa137fb5f8b1884dfdb80d8a236782bab0,705b03dc62ffee11f03e1a8db04102e23f19b942810ef564b8cc6a308bdf9c40,83984a57bb7a1c473bd9b81b37832555c7b6ad1321db082679ce2256efe7860f,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,14.285714285714299,14.3,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,chess-puzzles,Chess Puzzles,games,8ead28b45fcabd7e10f75793f4358b6a731e911ff7081633c9b12915bc9a0a74,8afe353633edecf228948037bc85836f5677d2d2e713f3f2ad195cad19db93cd,2299141735ffae39b475ae648b76af24afb1b26108db9220fd90ab92710ed553,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,66.0,66.0,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-13T13:00:23.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,arc-agi-1,ARC-AGI-1,abstraction,610de55f718144cf8532745e5d6a5088d56dd0cb69109ae30b2000212f9296b1,13f14ba6f4825aad82ca126924d0094b1510afbfda8816c7f792b70a3957e8ef,69f32bc8259b448a1f8af6930b6b12e08ba568889f8ec89626ba030f9e02c99a,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,95.5,95.5,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,swe-bench-verified,SWE-bench Verified,coding-agent,5f048c456c7548bd7699c944ec0fb313d336342297660a78002db123f84366da,6bed1ae144c6b5f3d03b5c3a48f78d9a27cdb80ff00b72393a2beabba1d256bc,aad3384a45b481d0d0f360c7ea6d46d4de011ffe30453fcf0e373637590dbf67,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,80.8,80.8,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,deepswe,DeepSWE,coding-agent,87e9bee21cbd1df04bc424abb64876f913627d00af5fe54fad1e698e57724c92,c2fe90d75526beb903835bbbe2ae4e5582d1bf9abdfa369143d517c93656c2b0,bfefa0da8b857f3d96ac81cd445976b8fee7a53a767175833268ee0896536293,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,65.2654867256637,65.3,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,frontiercode,FrontierCode 1.1 · Main,coding-agent,d4fe1c8f33157c73b7384f37392b1e5d77b9208a0f652490be895a500ba495a4,6229d7fced424b5210620af451ee0b9e52c7918ea14c5118d26ff63c8867f03e,e84a6924e739ee6966136bac08817f053eedd3e1dad6eb56cb049f8836dab90d,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,43.59,43.6,exact,canonical-source,profile_ad3d84b960dfc393e7214ac4,medium,medium,true,fixed_effort,,false,2026-10-28T21:25:01Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,ioi,IOI,coding,68f7f05338fc9503a5ac300d3b2a95f184c8f6e6079f03884c1d5c19ed10b919,400479054c6d76f314e255ca4d0fbee25834a0b02ea2db21024a87d990be02fa,a18b09073320ab85243b0e073208d91c90467926cd4a2d27a2b832c6764ada3d,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,67.833,67.8,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-21T19:24:33Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,livecodebench,LiveCodeBench,coding,14f3b6365cb65dc692c517b8856ee8f0735785b62cca89f799f58276df66a7d5,114e1539661087428751b773859563ad3e5b3b3a8536f80e7c08da73d995170e,4f3ce382986db9efd84e74b08136b5e4acdc7aeb74bef6c5244b43413bc19d91,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,88.652,88.7,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,scicode,SciCode,coding,2d604d8e8d63708ee89e0c93b22b273d7c19c1a109cb1c6cac04cae4e28c2d9a,25b7c166cc6a7e07bd65696d93d573261de529b94dbcf612dc51fa9a9850d146,a1957ec864f84f6d3231599a8a9b9cdfb930290573864e8857f74c2c4e1e5327,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,57.175925925925895,57.2,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,ale-bench,ALE-Bench,coding,729de20795f458d62b096931011a56b6c3ca12c60f2da1cdc333d0762eed4230,9250ec5711f956103cfbd06bc778457b358fe3e990a4551fd39f4f22b3bd2be5,f3311e4be89c1b73ae094ad063587aeffe70b07391548374dd81996c86f2f28c,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,904.3,904.30,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,vibe-code-bench,Vibe Code Bench,coding-agent,e7da7006a2318a5911c7d57255f4def268193612452c04599c46916545e62add,78b1f7042e6c0185a0fa14c7a80ac77093a19f652a6233b6241c0b72945d6566,0ba0f7ba05fb0ed5547d084d7efb554186ab22f9a38c91f0a0b9fbef5556b55e,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,70.395,70.4,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,webdev-arena,Text Arena · Coding,coding,b99c1a9abb802b08ad139fa1bd2a24fea80cd145aa9c25f718ca8833849ca814,98b9345a68cf987e361cce3c07146c29a435514f9c4639a2aad4785859fd92ea,5b2ccece8ffaa95d1b70052dfef45054d355ddad73d373b9b877bd66d5925eda,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1589.6728568868868,1590,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,mirrorcode,MirrorCode,coding-agent,309d44f0cba78432d503e431526c29ac2ff84134d4ec3194dc88a22cabb68db1,24c8769f9c20a07118cac446e55ca8810c6ffb7e6c5d71b26d36510ea787d45e,98e8b63d921ad13c340ebe3561a974acaa3554f644b6f9eaefc2a373a8938dd4,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,41.11111111111111,41.1,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-21T16:49:08.000Z,fresh,10,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,programbench,ProgramBench,coding-agent,0fc0567d03658f890ef88b16af7e625668785b27a9d9aadf0bab721a53f8bf06,38eb990fd2d70b364dde14e7a18c25aa254d25348c9414e0ff1ba0a4472c4b70,0b1207a652792e565e4ab5f2ce4b9bd7d4ac8ae5721eabefb1d59792fc7ecea5,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,f851d9782ebb319f6786ba101fb44ed4bb8c45939565232158e4e6dd17b1f23f,b69bd2958c1be3517037660ab77f5b9326a0a2d4f87a6b4f258230cc23432574,47b086322cea7a838c624123b6623395c0092c49a7fe2ab74a7f76e2c70f0b95,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,77.528,77.5,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,44085b168c2546e9ae206095ff7db51466af7dc6591a48e30a6a31f072e1da3d,d4a0d957f890f4325006867882a6e9c713e2ee763f4a6cf6d2979c42168b044f,9ffde311d7c0a3135f7ec2b8fd421177fd1e875120978c7e18bb3bc031243d99,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,12.121,12.1,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,sre-bench,SRE Bench,terminal-agent,5be21b9e576e98e9ee48fb380ec091dd38d56447a1bd06ff8cb9605b99696e6d,098e174e7723c268034aee931278b1c45be47b134254613454a4486e352d289c,228d0041f446b5a3d7e7d29c0bb03e27402bd7c00a644d0ad510d02f997555fa,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,4.58,4.6,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,c93541012439af3dbb243920686fda78156564bf24a12568340932f7f0daf187,2df23e2db8828eba3d20c28679b9c9098932807a4df5b562ff8143a533a5d61f,bdb882a998352194414c2aef9eef02eb80c32a1557da285459625023f8b85807,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,32.783505,32.8,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,apex-agents,APEX-Agents,tool-use,cab7bba3879260236a1d66f90726f73cfdfc5d8da424c222dcdf9532c98af064,39ba69565003326c245cfd35647b24cb380164f68fa426d031a284b6d1eda03d,e3016c190d81e28cabaa54e0616b06e6ceb9c11446f411ccd293d67b343e91af,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,67.8,67.8,exact,canonical-source,profile_e5060d9b2625a987949324dd,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,arena-agent,LMArena Agent Arena,multi-turn-agent,594174445699016ea937d6f2ce0a5678e4d7082701fd74ba418c157b7e742ded,470286f2cadc159e967bed47591ecd7c21def5d1ab26a6927c757df945e4f6e3,42de3fcbf0e58578f8438d70b04b1893870ab60a89a6d698f7c1be0d0d6244ba,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.013461022898167443,'-0.0135,exact,source-specific,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,112054e8ea799c29ba09a3e936dc91f730b52b34e675b8f7f38d5fa88d6cba3d,c969f0ff5a91549f26f6ea9358f4de608cc8563b3faec2f26c278cf4e8b8a555,93eb70ceba1cc7f4f7373a4267371005fac1040baafbbf53ba8621e278c071bf,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1402.91,1403,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
google/gemini-3-7-flash,Gemini 3.7 Flash,Google,closed,21,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,fe499ad83881d82fc01995cb9df65adb23ec904f4be26a74b3e7713f317553a6,560a7edfb9117250f2ef5c4af36d22e26979d177208604264bb7f59bc3008910,346f0b7bc9db3fdad5926b9d9058c64eb39e51e991b0b5f0b35f14b8529ea12b,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,7.143,7.1,exact,canonical-source,profile_45fbb04067e2dc09bef71f8c,high,high,true,fixed_effort,,false,2026-10-26T22:05:28Z,fresh,5,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,epoch-eci,Epoch Capabilities Index,capability-index,04854d1ff040a64addee5101187cc9f92075670d26f035c37082fa55484a407b,80128bf1fbeeab627a638a76179b828297636e823b8965825848fa94707564b2,5b121ef874ed48728f049425d72e52d7aca09a2b22b5530eae7cd4d9f740522a,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,157.98,157.98,exact,canonical-source,profile_d877af62a921b2b206ad3009,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,arena-text,Arena Text,preference,7460a6453f421448616021fccb04dd39f1c5e3c4e259dc3badaa3a70848d69ce,609f6433e57a85d1bc29aece0f235fe85e347e25ec12a0b2bf20074f9aa5e62c,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1482.4121421338182,1482,exact,canonical-source,profile_0d5bba5bc760b957bb971b9f,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,livebench,LiveBench,general,1a373f3db0795cda3afce37d63778467ec1459ea1f6e904bb96f5d8b8442dc79,3671599ae555646f0a9c97e7e8f3003afea6f41ba8764dbd274a64c5bf5acbd4,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,78.46089285714287,78.46,exact,canonical-source,profile_b41ea21204b5a7e098c39301,reported,unknown,,unknown,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,a9c5ffd6d7d563b6bb9a00209e472993f3eb101e58d3a1e654334065d3d4bf70,e2cc90045c78df263bcffda571d90a352f9f17b52cc0bcc99805e13e74407f84,16c4d8cdbec0efd3e09029bc345c875326c99fb0fe02090c1b765e95eb3285d4,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,43.04912,43.0,exact,canonical-source,profile_b6b0f9f89a407f75ea5e8b2a,reported,unknown,,unknown,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,simpleqa-verified,SimpleQA Verified,factuality,f7aa13d239827cebb31c111d0abf1627169556f2c0416fff58d68df9a50726b1,f67a0e3b4b9a471caca177a4211e2588f51ca771a10df8ecd58d688c1901be25,3f9353637d7d4fefaf798304b12dacc4c53ba0e42ea81f1284993764f419ef4d,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,45.800000000000004,45.8,exact,canonical-source,profile_9f288de94feeced4427eb208,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T19:30:55.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,gpqa-diamond,GPQA Diamond,science,92b4ffb0b9d26cd7edc52eb6a2d0f8954f506f4dfb5febaab785fb27cde8ca05,b493b3f7c7c40b773ca8fb90ad6862b107910affdaf6d80f1eb340bdec70dc52,7c268612c5c83f2d6bf0e679778d3273d8611ab34d7f7d5ea0ac9b3afc56f162,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,93.686,93.69,exact,canonical-source,profile_d8aafdf161c3cb55a509389e,reported,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,mmlu-pro,MMLU-Pro,knowledge,089709d07d950e4da395b05b90edfb9634c7f32001c5d8e5461fe71abfb66af2,54375fe6a46fa7840e3cdc0e28a0e5d5e4d255fd5ad7422837e12c83df7267ff,0b21cb8a5fb871fe1188b4cb0613a34c11f26970ba7a9ab001d219360477590a,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,88.602,88.6,exact,canonical-source,profile_c9dab4aa7fa2f362d4a64f77,reported,unknown,,unknown,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,95832a8dca438ec6b1008a8eba5cdccf5f7dce13ea76ce1b225148e1b01a9599,6aad45fa2e71a993b38cf69a5971e530b6afc3f269eed2a0151adeee30980956,e63e8660218051fe73ec5b9f865c4b6e4182286cf8945c04ba7347e139856fc4,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,99.44444444444444,99.4,exact,canonical-source,profile_9f288de94feeced4427eb208,xhigh,xhigh,true,fixed_effort,,false,2026-08-04T16:28:46.000Z,fresh,88,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,frontiermath,FrontierMath T1–3,mathematics,0aa9fccf4b68aef41a6d1b73fa634ab51859d63ef24710e273cc9bfed1f9a66c,996daf37bd8487671eb4001ab7bc11215b289c8dd9118e7ad129362b13250ac9,7c3324d48b24f2962ccf8492e8bb7abd941cae2adad1c1b4cda5aac279ed9ba5,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,74.73684210526315,74.7,exact,canonical-source,profile_9f288de94feeced4427eb208,xhigh,xhigh,true,fixed_effort,,false,2026-08-04T19:32:51.000Z,fresh,88,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,frontiermath-tier-4,FrontierMath T4,mathematics,bba89c7c3443c9e0c5373fa0c859c6d7d98562010703b30eafa735bd7d884392,15467be0f0b2838604e9164d845f830f5ee462293c0deca96d522701e7669038,994602caa30af06c383b9d649b7a2039606845c9ba9f3093dd42ace2159cb112,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,46.34146341463415,46.3,exact,canonical-source,profile_9f288de94feeced4427eb208,xhigh,xhigh,true,fixed_effort,,false,2026-08-04T19:32:51.000Z,fresh,88,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,mmmu-pro,MMMU-Pro,multimodal,8ac1f6bbc35f772917280a02cbed2c647dc600b35dae7ad4a9df2e34ee369ab5,6a3a02191ee03c67a7edb8f4b466705f5d5467abd4d58039d0d8d0652424f30b,b18c6e3f6d0c5344793766c2ab9defec2c5bf8a2bbd7159883f78828aed18739,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,88.035,88.0,exact,canonical-source,profile_1deffe7efba519db4b67e6e4,reported,unknown,,unknown,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,critpt,CritPt,science,1662997d5c6379f07228b112d4a41cb90df401523fc7625c46401bd6e2703041,2b65f6058a5b441123715c864e70fa37ed6067a3e568954a2a0bac2d1b71aa9c,8ae55df023be0e7f2e62d5823972ffd02ac3245bba85d29d76faeaedb4b39410,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,20.0,20.0,exact,canonical-source,profile_1718ae78cff960c44c0c9653,max,max,true,fixed_effort,,false,2026-09-02T21:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,chess-puzzles,Chess Puzzles,games,4643b669e36531b1cdf63f910dbd871bb348f1ae219c8cd5d122cd391d411004,d2c0c32a9b44d6917866a096ed2dadb93e83a7a6550cc64f91a4c761cf8672d2,da5c828788732614f309d013145d55b69ccc1bea1615c3284266d8bb0db9d3f2,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,46.0,46.0,exact,canonical-source,profile_9f288de94feeced4427eb208,xhigh,xhigh,true,fixed_effort,,false,2026-10-13T13:13:49.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,swe-bench-verified,SWE-bench Verified,coding-agent,acf295a70f88a54ecc55bd9fc570e516083783861bba601a4f8d1296756de6de,a47a33a73703130cf29462911e35fb47e2d415cb8beb2e12769dc270d37128db,b0f08309744aef993faee12e36e962b3149e5630a4dc4bf7c56ac9978392fd49,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,85.6,85.6,exact,canonical-source,profile_5f68ba2f45063ad59689c8b8,reported,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,deepswe,DeepSWE,coding-agent,dbfb894347faec92ba083ffdca70d7b1d28a2e973311086b346022a40449d2f7,9027b13ee5edc2c1d2b927df6d7cc4e6fd47c4cc4f9354b3322c513c60c50612,a7141b72f5abab97686747a8f456230d9df3fc1dfbefe0093c97e1efe26855df,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,57.46102449888642,57.5,exact,canonical-source,profile_9f288de94feeced4427eb208,xhigh,xhigh,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,ioi,IOI,coding,a5f0d0da357c09b2b7fbb8d235494534293aaae0e22b7ea7e1a38307401c67a5,b74f12dd219b2c7b4bc2f5096ba1b9b28a382f38ba221bd967c3e1370c5b96be,8cebbcab44039bfeed0638029e038c696cb0b4b26037af01003e8061836cf8b3,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,68.889,68.9,exact,canonical-source,profile_dfe941467107f74de26697a2,reported,unknown,,unknown,,false,2026-09-14T23:34:42Z,fresh,47,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,livecodebench,LiveCodeBench,coding,57cab98dcc9dd6df58101736eb9a5492f0e64df49526447e87a71fb52dd27f70,06e6076c44b9e1ce30034e63fc6e35d938720cff883d210ddb485c5ebdd5ec56,0deadf0164de399f3f8418b7c982fc6f516919472b3f8f7c081d190ae87f671b,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,87.852,87.9,exact,canonical-source,profile_a132c45f37998747339c62c9,reported,unknown,,unknown,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,scicode,SciCode,coding,1bba2f38a89b396c90cf2d9f7ef381fd49818ddaf7c0b496bb4e5beddb03ebb0,5fcd3efa45bdaca870a22d1a6f225e605f388afe60532e08b0be369abe0415cd,a8c79cc3c4c00a0e4d283760bea23df2d04d38daa2bc97bc3c2fa3fc632b6df8,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,53.2407407407407,53.2,exact,canonical-source,profile_1718ae78cff960c44c0c9653,max,max,true,fixed_effort,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,vibe-code-bench,Vibe Code Bench,coding-agent,10c46a8ead36a8dac641899e291304a2f440db86124feb7feaed26d6e9cfe298,c651f21f6a2113d88db9c9ffa17c8bc355a9b168eb4de63ff16e0d771cd13217,f1408e393d0b8c73582634616b65a1fd2f2e996e2ce024bdcd8bab6b8c5d794b,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,64.696,64.7,exact,canonical-source,profile_2fef4e63c947b0e5bf3ecf53,reported,unknown,,unknown,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,webdev-arena,Text Arena · Coding,coding,0e69ac5614d9f8822e2bf0f8a7cd41adc92ae1b8202a4b82ab12d5217c336403,13fb00d0dacd44a6370183d7ce8b842943f6c62ba9147d4c0c7d9bf8c78225d1,528c2391cc358fca591e32e44ce271af10ce34b174043108968d55c4fc07ec51,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1671.885286248237,1672,exact,canonical-source,profile_a6cfdc05f4f1b956853cc974,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,mirrorcode,MirrorCode,coding-agent,e2c936dfd4f9ebf123e2a73c3c9d7600fd77c458a27ce2e3d95620e3d479b345,d68639702a82412f3b1e6fdf3070772f7e02c3360d38dccdfb83cca41806c890,ddd08864e99511b58aea5315dffaae5847523b2ab61a60c7f7407b498c073429,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,52.22222222222223,52.2,exact,canonical-source,profile_9f288de94feeced4427eb208,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T16:48:06.000Z,fresh,10,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,programbench,ProgramBench,coding-agent,78e83cac0b065d0edd948187d9680e1ded773b771e9427c6d09b7682cd84f392,5769502f4107dacf0a4d1892f931429e6b6e675f150a56beeba6d09a1963de0c,46bfb827e85a5568bbe27fdb40973fc208ac66ea68de655ff415c90750dc99dc,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_89bf2183d1e84ef4c1347f7d,reported,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,5f388ee675717ece7bb3a1d5f82a548f5abb689c00afeecb3e60401553ab76e8,9174d1a1997e8cb74ee390c1e042ba9dc1418469cc6560011ed02b416779e929,0351bea91b7a27ca232dced3bc5dc62cbb4600b62487dc632833f157f0ddcd3d,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,67.416,67.4,exact,canonical-source,profile_dad7b63bac0ecf821f75dc11,reported,unknown,,unknown,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,b68803341ab4e279508ecc6359b714c2846c1174b1e8c683d0906b9415eb7a58,2fbe74097c7d197e5fbbc7146175b08a28afc8cb25b048c794ddb1840414d50a,4cdfecf985107137acedd858b42df018a582dfdfe0e50e105788591c82bec0e4,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,34.343,34.3,exact,canonical-source,profile_4c799730d8ed94b1c16361fb,reported,unknown,,unknown,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,e96870b07bc4c4d14b7fcc3503067f80997f3ae2d05102904fe8d54f5ce414e7,74f3788b22416d2ee905cc0d46c5e3b6b9ec73fff2b2a9726d7c9971c0e5ce18,29ee69f49dde555d2727d00f373c0bd81259e7c9bd82007665bd401de1d6542e,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,51.340206,51.3,exact,canonical-source,profile_b6b0f9f89a407f75ea5e8b2a,reported,unknown,,unknown,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,skillsbench,SkillsBench v1.1,tool-use,5849cc4cc83c700fb7dfd173959c53d369bd6504526768d924bcb8a67384cb30,a2c923d59a904bd16ce46fb77f8585d4df9d2b251f29f53be9cc730557b5847c,5a87a249627548ba41ab1e22adf8f465931b323f1bc4cc1e11b5f7988d5b1a3c,release_ff4a6e288c3c17d588478b46,SkillsBench · Qwen comparison run,qwen-vendor-comparison,pass_rate,percent,true,70.2,70.2,other-supplemental,supplemental-source,profile_646bb68a28e877c1466e9d75,reported,unknown,,unknown,,false,2026-08-12,fresh,81,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-8-a95b-card,Qwen3.8-2.4T-A95B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,apex-agents,APEX-Agents,tool-use,a62e1a674f8346616c6b8d295d2debe3212ef032a1c1ac2e5fe98632d4258817,c14a09655774f59dc545f7886f4f2f8a0c40221ce2f2f2020b67880cb9b0b3d3,69c1d7d460388b7e24674149491f6ab3cc4ae90f0f53207a3102d10de3d6409c,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,63.3,63.3,exact,canonical-source,profile_1718ae78cff960c44c0c9653,max,max,true,fixed_effort,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,arena-agent,LMArena Agent Arena,multi-turn-agent,b236ba8fc2b1a89c26c09433504365488342750717bc5271a16473bc42ef32a6,0677237f30807d75c2da925c279c1f0ddce6d3e8140612b04af4e72d2faae0a8,09dbbfd1bda806f47a854e3ac7c53c0813923a4cc900df9730436ba9963a486a,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.004482244232743595,0.0045,exact,source-specific,profile_b609cc1047aa681b2607f1ec,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,e0c218e6fb00729ff3368d15873f55a43699bfdf77221db167ce530d089a8a32,6452fe7a7931c77d667dba149ae1669d4752c5d8be7b375a0fc0902a0f01dc4e,a1c5d9828d35cccaf99bfed5c00522b98c774b8b126b7cad63a4f827167cdd8a,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1611.45,1611,exact,canonical-source,profile_b6b0f9f89a407f75ea5e8b2a,reported,unknown,,unknown,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,6c71029faaa0ae57fed15e421b4336fc86f97f5f4abdd7adb3d2d828eff46316,444da58c91a88855a8c0deb7f48f271df46a7457922e8bdf7e9d4621e9781433,d24a96efda541fed6ee85c3d0749d022a0a3a3e770a8b96ba60d65bdb1d15153,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,12.857,12.9,exact,canonical-source,profile_466a5cfef0b0a57e549962db,reported,unknown,,unknown,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
alibaba/qwen3-8-max,Qwen 3.8 Max,Alibaba,closed,22,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,1aa16913a61d676e6b9f8c76e9e2506bab57f3d4a591da2ca49b69238640cb1a,9d815173a6e2f7e9a1ec6f0ba271399a44ad3d9b3cd57853424bc75cad527bca,ed98c8443e3b747933fc9595ca5529c72ec372d970230b74353a8645611b5d55,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,55.154639175257735,55.2,exact,canonical-source,profile_9f288de94feeced4427eb208,xhigh,xhigh,true,fixed_effort,,false,2026-08-03,fresh,90,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,epoch-eci,Epoch Capabilities Index,capability-index,ce56ba357795f082584e3bd31969a228b7bf42512e02331a1565dd09d8694d8f,bff7bab6825f85674d1abb36aaf1b76b9466bf650be4b7d1b6b828ada475cb1b,b35cb1b50629a6306d4df1f211a6fea82770620aaeb92502763c075c1134f73e,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,157.64,157.64,exact,canonical-source,profile_e5abce6ad7b99a4c5dd13f88,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,6f7f230d692c8fcb0306980b22da27c56fae0834d4963689f2c1720e9d52c7a0,423d75f4c16815c6ddcd3f1e86e42fc6df053ae81e167781ed1124f7a0f6882d,a456e0b71d795210e1107460a5a9ce59ce52b3d0a32670c5cdfc96e5d55739c4,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,30.546812,30.5,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,arena-text,Arena Text,preference,9baf7bd7b9b206c75cb4c0377e5975d1bcd936a9b139183346b31893c93df1f4,800447b476fe890ce244fd8a666391f80d2c2e1071637172582f331a3ec4bc9b,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1495.9450931791241,1496,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,livebench,LiveBench,general,ce40800b95539d3cc054c280f26d430b8f6b47f7084b2ea8dfd62c192c1164f6,10f5a1b749deb65e6a21343eb1dfca56922166bbd62d6c17b19e398811252200,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,75.82852380952382,75.83,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-02T19:01:19Z,fresh,59,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,5e79ad89b115ee70a768360291082b0156bd4574fce6bd56e94bb0a9cbd2027b,7cd33ef81063f6764e32fce6e79eda6adedade19c6d47c0732b02d87a8ffdaf5,7ffcad33b4c800c8c032dffb42979c558f328976374c4179f10e782bb08fb9b2,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,44.52,44.52,exact,canonical-source,profile_491d6d7d0d5b0ccd98ec2ae4,reported,unknown,,unknown,,false,2026-09-09T19:01:48Z,fresh,52,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,72d247f16eacd8431ca6808042c6969a28e3c5f1adabdfb7ba6c253fe56dd122,7a9fad8f5a6f750b218e91c5cfcdc9a0daf8595ae62297c2e3db01a34f2367f7,0d2c4161fd21057ae8cf2e5ae5ee148326e3b0e3135eb8a06a3b8b9b5f989ea8,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,47.822057,47.8,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,simpleqa-verified,SimpleQA Verified,factuality,815a62bf70adf8ba87b4e46333c767a5841861100e22aae0fd99538a6260092e,81b0bb66980faa473aab681bb2dcbf7abc65fc2c98047a1c88cc5bbf7ee2291b,addbe5e3a24d08a8962ffa2f57bb6b6881faef74d4cde0934b7d57db69e9d41a,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,69.69999999999999,69.7,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-02T17:15:36.000Z,fresh,59,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,gpqa-diamond,GPQA Diamond,science,9a8061a3a1b4af2e9173e7f60f9f9aa31c4b8971588992065d56b4952925b360,55ada339b371c7c74edf56f40a4726b14690f43ce6ea53fc7ea9c5f6e931f05b,f9dbd2529835217eb544ff449b521be800034aff9e3f0dfefbae323c543c442c,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,94.444,94.44,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-02T19:01:46Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,mmlu-pro,MMLU-Pro,knowledge,e92e0517107a1d5403a6757754ad9b12cb493f05dcdb0a9017b9a4b1b0ca0e70,973b1c93671fad29bbc938c594a5796219b376d44ea122edaabb8b50846377e2,ced96f8760d6bef29d512a74852888e06ddcfeb5faa5c459628c337f13667018,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,90.221,90.2,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-02T19:01:55Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,9e968784e0e19a3dfc8afbcf898385171b68d4c19d290782238254ba96a00391,dd44165ebefaa49b7788de913da0e6fa2f7b7e2cf5de66918bb04cb034e0decb,b42224bdd5801441ea632d9fc7232b935c6196270e058988ee5a4bffa469c6e4,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,98.88888888888889,98.9,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-02T17:15:36.000Z,fresh,59,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,frontiermath,FrontierMath T1–3,mathematics,8560cde5d38851de3905fafb44661198d8c475201cdfc4b523d05fb9f41ad27a,58240594590e36e7d13b65be12128fcf98cb0b93b9cd54a64875d13d7875f405,fd8b819ddb1e17afd50e047bcc1b8c90c78e246dca57235b7fcbef49bd394df9,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,68.42105263157895,68.4,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-02T17:15:36.000Z,fresh,59,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,arc-agi-2,ARC-AGI-2,abstraction,48431d219c6d7971ec81e59fae72a7117a3439271d21f58519a824d9e965d317,957931474cb6458de726e16f61d5c8c69855bdfc4b8ed642e01125595ef65466,4adc9966b4f2a764c90a6ac1a6cf7f028c814f9d7576eabe5c7198b7d98f41c8,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,89.16666666666667,89,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-30T18:53:20Z,fresh,31,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,frontiermath-tier-4,FrontierMath T4,mathematics,bfb8bf3d0315f5c77b2de0afc4c4bfc5a1358d3c8cc796c7527a2518607a7bf3,05be8d381c7c72ef2344e4a734fdc7542a4d7a88124af82e7aad038f7707d51f,bb63df4a7cc1a4736fced56d574fdc35e3518f8449aca663fff044b6db3d874a,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,21.951219512195124,22.0,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-02T17:15:36.000Z,fresh,59,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,mmmu-pro,MMMU-Pro,multimodal,af7c015b76397e930dc220e7d1e8794ab0057f4f5f943c8b21b92733af83b7b2,39d80cf64ff7626d6459bb1a597d7da158c12a96b35d15574b557880d6d678df,b28886a06155ebea2908997dca972f96c6ac0851f12bca6109b614d63993176d,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,89.075,89.1,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,critpt,CritPt,science,9b9406005cbec6f7edd3ea2b48cd5fd2ce9d42d00f0dd500f2a4414d209c7431,e998643de2b3f5f01c488fc53d411e939727800ab35994922defd91ce122d66e,db92ffb6647156a3f7a53f013696ffabbd5e467bfccdb9fd1169caafb8462bbc,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,18.2857142857143,18.3,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,chess-puzzles,Chess Puzzles,games,53c6396e58a7d13ffa576e4b89c351b102d87888c7e9326f302c177918ad753c,56935ff99f09ab44f32ec42ccd0d4abdf5a98ae3d617c7b043317381a3186352,86392548d9c7be0dd49e24dccf71a143090ca81ca435cb34f7d75388d3c6b430,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,57.99999999999999,58.0,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-13T12:39:05.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,arc-agi-1,ARC-AGI-1,abstraction,c22c92b8df5a490f45a7b4cb6909328d677abc7478ee1cf47d3a000e5bfdd6d9,cb896567d6622bfc7251db9ce619afc49ab243d3aa8d32c0827d2f61572b0245,17db96e23ddf00b6e41df1eea85980d0ccabbc8c9964e066d35e3272a5fa7a38,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,98.5,98.5,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-08T22:23:14Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,swe-bench-verified,SWE-bench Verified,coding-agent,e3e3082b41054ee38b51f80e7c759d8994c6abbef4133bc8c7b4b4122fa554e8,4361c8c54c29cb2474622e1417a28e3afd235da84b62de4b1d195b50964c517f,e418c746c9fd4e8e79df63f385cc082729673fb65ef8b112edfcd4f7ad090bc9,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,80.0,80.0,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-02T19:01:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,deepswe,DeepSWE,coding-agent,0ef64a9deb630477bdb9934a18e2c80865dcbb13eff92446b3e52ae56c9edf2d,f991d213bad7176a6025690307b002e94cf1fc3203f20c047473105c96d32976,bb706fd403a809640e72b3480dd7b01fc30d05d6a6ef39a0eca2982aad7ea2f5,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,73.8255033557047,73.8,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-02T19:00:09Z,fresh,59,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,d218429b572610024d0e25836ffd7391b1c6b6402de7b9f79fdeb0bc4c84802c,43d5a0afa3330719a5906379ffb97ee8f4871bccf45c6f0c9db4a4562f727593,1a3f04128ab3f50c15d644fd92e81515a8ba05cd15d9110822d97bcc9b122b07,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,94.86,94.9,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-07T01:44:58Z,fresh,25,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,cursorbench,CursorBench,coding-agent,32fd2fec306edb96ea5bb3025a1bd4f41ab097b1b20aa0110e4676441a51fb8f,32a0f9048717254aef938e84da5409c858531c2fef97e760dada72efa0b659dd,8e4831fcb7f1529b499ad06448224021e0de9d56f07a8d4b6d660034f45e079d,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,39.6,39.6,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,frontiercode,FrontierCode 1.1 · Main,coding-agent,8131ec43c04f482ee5ba4f68a85ee3dc18f19e7b965d970edac1a8c3178bd184,415a858cb747cef45281a0f7ec73af0a9aee4cdf8c02acca28c09ab85009c55e,e13f7845f4e5319ac18544de8a068a724cb9f1029b61ea5c40bfe7a7a860dd87,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,41.19,41.2,exact,canonical-source,profile_eb17f29c7f8397dab70e09d7,medium,medium,true,fixed_effort,,false,2026-10-28T21:25:01Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,ioi,IOI,coding,1919524c2c4a77f08bd9d2c0bd7f9e0f4d47b24e14bac372f5b816315b3bd1f8,91c5ee7dfd069ecb8cbbcf68c42021ca1516bbb89e9160c396d225a86d69bfa1,f17724c3bb48d0519fc2fd15feeb5705b2cbfddd9b5b7f6949c1ecee799fc396,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,56.944,56.9,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-21T19:24:33Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,livecodebench,LiveCodeBench,coding,ddab518472d9e83901b937db72475625c1544beb712e628770ffdcef18488efd,68ce26c4a46f204b4815df0e911bf053488852ef3782352ad2147c3ac8679c5a,e1ef6b16ab215600262ec2781d984365cf2d293c83904b7214d797c5ce9e6625,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,89.484,89.5,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-02T19:01:50Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,scicode,SciCode,coding,792d38aecb3226b480a4dd3fb9538c168116026b2d37487294effee210dd1193,5944950d05df568b58d18046b800cf570afa688e97dd2a0b1b2754c0e6c0a13c,0430452b7e8bc355286326ef670886234ae3e58c8aed15ecdf1b44bb26deea9c,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,56.5972222222222,56.6,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,ale-bench,ALE-Bench,coding,c11e03d51037548366a630fdbb47629fc69699d98c8b174e926f9a6a3f53285b,c5eea9c76c004d490a3ee87e13bd65573688ec82ba47d3ef26b04b7e6530858e,4f3bf61e93663a340c7327d10cf822d76f59e596b705dc5911fcda0e91cf97e7,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1269.72,1269.72,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-30T18:53:16Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,vibe-code-bench,Vibe Code Bench,coding-agent,7fdccf72fca10e02341cf2e0966bd71d68ff65dc18a0de8e162904ac8a581365,b0554fb7d1afbac6c13e7c11047ce0ace713feed28fd160929747ef71b0cebde,c0979d1c41a9b6329d8820128a77d592ca24e49aa0fb7f696fbf3cc1e6466f11,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,78.651,78.7,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,webdev-arena,Text Arena · Coding,coding,6ce3332183e90b8568918a0cf2ee8031993b11b71727d9ab6d6afad5da5e09a0,cdc4d4fe95b920ea1ee3ae0fdf1239b05f11adcc90455aa23af6754086045f41,a04234ca002d2adff74fbdfec3209af34f7efe4323facfb4546a0037dda2a2ba,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1583.2994176433617,1583,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,mirrorcode,MirrorCode,coding-agent,ac1ce81702d2261540ab88fa83b0081c9f9cf16a05721992cc41bc19d0177ea6,63df43edd2b51a3fddf01283ee32358ee8b4bc789896b1d50c0b20a6c4a483c6,80921b9acb99c020d9f8a8e498bf0cb8d2e96147c447722eb4b9c3345f038061,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,38.88888888888889,38.9,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-09-02T17:23:38.000Z,fresh,59,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,programbench,ProgramBench,coding-agent,ada3457d22191d16e035a09c8d61ccbd420546d2a42e306530d3bf3b98d917dc,5655c1785e4564f276016c7eac6c443eef1a4d678bae0c7c69b3e79f508f0ee0,fce468e50c91a38d05e6a0489790c14462f16e1c2d2f95598199a0cec501a7f7,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,1.0,1.0,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,weirdml-v3,WeirdML v3,coding-agent,6aaa24a51270a2b1eabe94fdfd22ce359568de6a96fdcc8105568bf5ebbd6834,9b5f28438d911994bfc94d3d8fbd434c6f8a76750ce3649dbc814aecf10b5064,05f1b4b7e2df102c6c151f2f369492684f7e9204759e7d61eb444470c95a105a,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,7.378749090909091,7.4,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,649f040d7d44909fa0dc706cae401dea6a92a0ea5b731ab95aa3f9540362a966,7a0e8c97cad0eacd1be7d2bdbfecbfe86c62993ce3fe425889cbfa3ced747574,b6260bfa498c9c9aa873df3331d3a5a7f0bd99b50f8145c96ab8f9b57d28e262,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,81.273,81.3,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,a6a0bda4ed7ff6cb065249cf428e830e5584235204a34098e8d4caf7cd1692b9,1c5aa400b0180b2efbe425305990eb1ac3c36852484996e5613a462bb5ca4c62,156ae98d4ed1eeddbb768c263150f414c887cc09625e11b2003c44390d1a28f0,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,19.697,19.7,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,sre-bench,SRE Bench,terminal-agent,c3d89238e8102dbf6f2c841af46dfd8c328c3df9ac3c2f86504e3d230c9d1b15,d6ca4420150911bd22bfc68c5270dcc92ffd9184cd1559a1a21cd68d8655fd4e,2f4cb7a86dd755e3a7c0ce47fd7ff3c19ba06e8cedb681d78fcfae49724579b7,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,11.069,11.1,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,dcee4e224d8b350c452a2a1ea72bca5c4ce683d934dae973d4f1095bf2b49c7a,62f133a22b477e9791702bd70689805075a0b7b174ea9a795c4a61240d77954b,4235fc7207c75ee04a543d4681312f3f8efade84a225648c6214468fe9fd6aad,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,44.948454,44.9,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,apex-agents,APEX-Agents,tool-use,64a03181fd9388cb544c8ce58c8a9e76c45a974aa69cea60c35ae1b3557520ed,218aaa355128f79980e1f008a3d87f5cb4dc678b9273259e4e3256a1fc4569cb,d1db83c8635ab08e0fd6f42727ac51e27926360388397753feba6a0c62b75375,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,64.3,64.3,exact,canonical-source,profile_51f85ec2f1c37d020adb2824,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,arena-agent,LMArena Agent Arena,multi-turn-agent,064240a7b13a958c650b93bd79e2b5fb9749133803215982b77e9973bef45f12,a26abb6893d940023cda5ee16e3cf73cbbda5f993af823af4afeeb99c1c3e263,b073c19fb3796e4d64756f6f913a7498a97ddf1bee81a4ec61daee204a62d9f0,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.026495450925865173,0.0265,exact,source-specific,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,vending-bench-2,Vending-Bench 2,multi-turn-agent,23a0ec63a24a1b8745b9e1b4c957cd828f15f561a23ec602f55db3be766420f9,587858129d9a12a342dcc4d19abca899c72df47653432c81f808317d36957a94,97b9872436953baf92d74f92534fe81f89f3cce8d3f8233af47bbf14ef27df99,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,5093.793333333332,5093.79,exact,canonical-source,profile_cdf54735b765147ceab18bce,reported,unknown,,unknown,,false,2026-09-30T18:53:16Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,21e05b7a97583cdc9c9cc8b16b4373f92f101c0f01e210cd5d3f2a72f436c80b,c62b9a4e1a3cef0c3c472ca5753a109b240e9abb32e876fa3af54f6e1fc2478a,b063799c2fbadd40a0364ea59f31bd847cbbc1211136c30d0ea42ed4dfb677ef,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1445.43,1445,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,a7889fa105f8a1a96ad0e52bf51278b55a3eae85a8e78da136c8c65e33cb11bf,2f028da222dd86a41d55ce855ac03f2c0b3072fee82955148528b0c43e94db64,8b6d70e96179336c971a9c15e7d86e28b71f07e2bc902b6379552a2a99e62e9a,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,8.571,8.6,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-26T22:05:28Z,fresh,5,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
google/gemini-3-8-flash,Gemini 3.8 Flash,Google,closed,23,weirdml,WeirdML,coding,6bf282189000dae2fb90e20fd195e6f6f2ff908cec3a1c848dd879295899d666,632052bc13c415ad922081735ca77983308e8abfef8c388c4466f95df28f7c7e,3f9527302727eff742bf8f86a94d09a94a1c2ef8b6eb4e56d7455511ee7ff882,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,84.81,84.8,exact,canonical-source,profile_6307c6a5c62851b4f5c8d080,high,high,true,fixed_effort,,false,2026-10-08T22:23:14Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,epoch-eci,Epoch Capabilities Index,capability-index,52c654855f194326e2a6abe3fd8a2552ee26abb079f2397d65bcbca487adc8ab,e0368d5f4f87c42f15ad6cc07e559d4269386e35693372c73a1c7d79458b5c20,12ffdcd126c0012251bf3a62e3b8bf5363a2b3adaaad2f1f8e42a37fb06f5901,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,157.37,157.37,exact,canonical-source,profile_0af9228b335da865d7243f2e,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,arena-text,Arena Text,preference,0d1e23539ba9854f771633c91f89e306f8cfd261e5fe56d13cb23c3fedcffde5,aacf7e6fd53371bfaaada37d3764500a134a3743acdab54ac504ad24259567ef,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1495.0456185334651,1495,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,livebench,LiveBench,general,a18d0ec1d1ce1dbb4e1620e5393cea5efc09844424234cdcf6bec176e14b4bf3,5231292c3314541e269f278dad0aa3f3bca5ec1b68ad1779918f20e47b44bc4e,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,81.58864285714286,81.59,exact,canonical-source,profile_78f4b55ed67e2456efcd3bf7,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T23:43:32Z,fresh,59,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,6e1cb29141d825bd6bcec21c0c16c7fc9c5f86647dd2052dd28a9f58421f5f96,0dce3b5b6fdcaacdfe636c10ae5169d373622a11dc81c28cbe763f22db7ab062,d21ce42d35c7bf3d753ae910b9a95833f4ea328b9c3c09d906855b4d95d290dc,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,48.702502,48.7,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,gpqa-diamond,GPQA Diamond,science,91d2433b6ae5da6cec1f9a3035ce5a89c8aa609a7b54f084e3cddfa94e118a9c,88a3216f0bd124e0bbbe276c53ade1f0a1ab894648df2aa9b2e19f2c89162ffa,c6cd16978a52d1fdf8289811fc0a789342ecc7c8bf835f4085d77ac568a3246c,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,93.535354,93.5,other-supplemental,supplemental-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,5b135bc4beea9212489cd83bbd4ad3d8f1f7291c759ad44c3df59b9775af7c8f,d882217a9a0182bb2369b720bec44fb60a040c9cc091ca9273ae6e4bbe288126,33ceb2e0d810bdbafc77f518cba2ee9be1b6c409c51b20f9b5f61314928fff38,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,99.16666666666667,99.2,exact,canonical-source,profile_78f4b55ed67e2456efcd3bf7,xhigh,xhigh,true,fixed_effort,,false,2026-09-17T13:30:27.000Z,fresh,44,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,frontiermath,FrontierMath T1–3,mathematics,b3389729fd1736940b7c2274391211d137c84405fd0054c106f797739f2e42f0,d69f81477ca97cfd6f7138d96f5233b350fa7ced44de95206ff6351eb0be9021,86c9feb2114dca3e6138829322e72fa076845a3cc3f734638a7c5460af8e3d54,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,74.03508771929825,74.0,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-09-18T16:34:54.000Z,fresh,43,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,frontiermath-tier-4,FrontierMath T4,mathematics,7e7549405a9339a1999f97c9cf2915f382920af91948cebc7f146860d6b1d2d1,f966d5dbfff7f2ab08413c6b1c14e437c7100d892ddd1a516d9a95a6f8877be6,e0c81ff631e2766c70cd005a9495ad590823b345eb9f683bc4bd3141d697ea7d,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,46.34146341463415,46.3,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-09-18T16:34:54.000Z,fresh,43,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,mmmu-pro,MMMU-Pro,multimodal,f62902f247c796ed55317b94f0dc6cdf8d8b3b509c945dc2fd8a9197f8e000bd,d14611516dd3da9f3d9c19a944708be81210e25404adcc2d58a1f73f6a8a1819,35208c66fbeab756a22596cbcc77f6a4dcaa0eadc438cb67e4c75620661adfeb,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,82.023121,82.0,other-supplemental,supplemental-source,profile_78f4b55ed67e2456efcd3bf7,xhigh,xhigh,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,critpt,CritPt,science,2295e1731660cbb7309aebcb08426b3b93ad74b8805eda36932504b4075bb79d,559e83a6bb6a171b657e908173e5851521c702f2cbcc499a3a65c87b9a367a9b,b55d42e298f853896bd90e96318536c810e5ef1b11c5a0011741a4c8873f913d,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,24.8571428571429,24.9,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,chess-puzzles,Chess Puzzles,games,b7204a756d3a9b3d63fea3612bb18279a7ab9361bc03ae9dc07e1f924b2bdc39,5e780804a6a1c350006bc56fd629a215b41580214b8917fe796c1e86be4e3b1b,698ca93537a6b601724ce4496086bea739fec21a751e296a8469e2a2b9e537ba,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,37.0,37.0,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-13T13:00:28.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,cursorbench,CursorBench,coding-agent,b550f8a3ba550c58f623dbde49d424e0a6ca9238be853a682b81f14c247f5a7b,ae0d1ba853339780fdfa71563fec850aa1e6582692a820dba46b51777e5f3c70,85242df049ef2b8b8c92f012e1abb8da0fc270a93b0880d2d566255bdc48607d,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,41.6,41.6,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,ioi,IOI,coding,0d77ff9a9623b58142d669317f18f01813bab0c55cfecf51e93df133870c100b,1d719a9afc7b43bb172ed68ee1b7d1c9c14e3e967aa3ebac5485beabb44544fa,b73cb9cfd5772c0b4423de958b9d9b06ebe958e61dc8b01dc63fd59a05594718,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,56.556,56.6,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,scicode,SciCode,coding,95202aae34082aa5e748736c63ca1969e6e58dd756aa7759e0a7af763cf1e3e1,e62696ad1467f31a11096015cad40f1abf8bf87484b528691672d98162012264,58a64f5e9ce1fef55acfe29433e82b2f2faaa2005eedad9401268cd075663f6c,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,58.796296296296305,58.8,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,ale-bench,ALE-Bench,coding,2bb027b76ca04e8d629283947dc81da4961264e9699432278533bedbffe6b250,d7829a34a125b4c19450a08c957b580e68b16730433d1b29136b348bd9335b07,75eaa8f3b1327dbc8faf02c4746ef0fbbdc707a0bcd4b9f0f58effcfe62d76c1,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1688.35,1688.35,exact,canonical-source,profile_78f4b55ed67e2456efcd3bf7,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:15Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,vibe-code-bench,Vibe Code Bench,coding-agent,ce6de1d22bcb970844a24272f0beb852dfd06d6d884c5973e42e260ebfa1266f,277ae38969478f7dfd43c66e9c52148ed9113b5ccef6fe5c3bf9b6b06aec5d2b,ed6e601f1834b4186b1e93aaa9189efc8c7d96669573db6cc9c9fa45d7bf783b,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,85.856,85.9,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,webdev-arena,Text Arena · Coding,coding,7719f84c257eda37adc39be2286d9b37ca5c18ffe1122e78ccf40cfb59a89d48,33864ec695e1af57b78c9606275b09334a91d63f4c536d2a098eff4b0bc20e19,c8f24338983fef57f4e349cc387b6bbc685ac247a220cb574337e0ad59fbcff0,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1656.2348593321256,1656,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,programbench,ProgramBench,coding-agent,ad953d5b0977a96fe7b35ac1aaf5b6183a6f34d773fee8d90199ca880402d639,121d627c7e0dfb78a17972b0b4902674344b45d1fa15f35f2709f275d19eca84,5df792403d4b41930209ac907b58def1ab298e26079a1aa3d4a3bb2b9b408710,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,2.5,2.5,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,weirdml-v3,WeirdML v3,coding-agent,e81a9c8a7fb1df45cde4c8f01a8ea81c691174d1d8af133609ea0b5ee80b0d77,f63f2acb5dd9a3d3b128298ff50b07a23894f4c2c8143d8566be96ed7361aaa9,4bcb0820c903884cef58ffa6b0e2e0561d1236a7edcf7885db45c8387083cfc7,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,6.65496840909091,6.7,exact,canonical-source,profile_78f4b55ed67e2456efcd3bf7,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,1ebdf57d07abe9356c64a5e1df491dac138da9cf888407a039d70c87e6e6277b,e6f54626009d0ee547bdc7d45a334b4be4a07a26afd8d4ed58774f2e74a9f404,d9658cff2811bccf68b9e4b57ebcd2bd82735d1c94c3acf02c2469da836376ee,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,79.026,79.0,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,b936684823ea05ab2a4b3b5d06861925cd601ebed22b74b8b86db70cd6b8b4a1,0b76860469e426fe1f70ac4f67c0fd0feda772d955f45c15f0cb3cf832c7806e,3f83a099e0def8f132d6683e45904e47eb86826819dd268aa71ac7896154e759,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,24.747,24.7,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,sre-bench,SRE Bench,terminal-agent,8fedec5b0d3871db915178db3c72ca16bbb6b0c7295c7edc8323a678c754fbe0,342a2ed27fbd79bc10dd87d33338ccdf743be9941ed77bb8a546556aeea5c5e9,ebbe25af433b0875bac75018c458896a822255365f8166891193542b453ef192,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,3.817,3.8,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,f9b9ac3bf45d7a727ce0f71a6aa16ace2c6d1244f9642e8d03c8160ab790176a,50c00ac134b7a95b5ad9297725a664a70ac6647a3b215f3d638ce4c4c0e559bd,adbc603baf07b6f79988c40179f4f48058e375ff7e6dbfc32b6f7e42de321d62,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,50.515464,50.5,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,apex-agents,APEX-Agents,tool-use,b71a4e174a3dec3509b89841136c2afc4ac691872a9da7e5dc3d8d3a9751f9bb,9c246ad76719a109c561e59999641d0ad7033b2c0502048f62f2411ee0835c76,6aadb29e85eebf458048f55a590ccc9fa76c1dc147d10d20846944209f9b0a7e,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,47.6,47.6,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,arena-agent,LMArena Agent Arena,multi-turn-agent,a6ec8c3d77e3a16ad86378f5150ff1b425c0e139334db25084b0ef07366e7727,bba25f4335124747ed80c315c5108033216c6bbfe5d92b6a06263b2f6547a54b,f68c48e1bf9af85e5a4b5469ffc59c4e300fd54024f5c72e96f7b077bfd95ef0,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.02593617325354452,0.0259,exact,source-specific,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,f69148e126e31db678b1c96bd10a068da927100c7689ab158eae1b0d7437b7a7,9fc30eec6a043378624ecfec2393a146d1223c908dba7ac3df4b86320d5b6b29,a87435ac4b12021f4f45aaaa031ef9013ec6da5155bee6e5f702e647e498850f,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1680.99,1681,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-09T04:29:27Z,fresh,22,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
meta/muse-spark-1-3,Muse Spark 1.3,Meta,closed,24,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,1e45a73002227793243d0403e274d8dfd6b7345a0fb28182f3727c34bac38c9b,fb5440443a1ddda247faeeea42cc5cd25d5685c9c9c7befd2ddbfac8e3799d6d,f2455f9c64ddb394079d81288f0a4230b467995199f5f3beac580221cb2dc37c,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,10.0,10.0,exact,canonical-source,profile_942867dc5166c18b24f5834d,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,epoch-eci,Epoch Capabilities Index,capability-index,71002ed25ec5d0e9ab694c8231d7cc7d5068ae568915ce7de78c7f09664d9af3,72c5e65dee5dd25fbd235ed97508bd1b6cf4641436a1a92f987d6dda53ad0acc,49879babe65dbd853ef21b681c6ef094aad3a806a8c870a30c07a97d5222951e,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,157.14,157.14,exact,canonical-source,profile_54ae75f31e2b8962dbf99cae,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,arena-text,Arena Text,preference,28d92b17307aa2678fdf87487cd357d683b4b8c3708b237d0ce02bf78677cc94,07727ab00a9117371d4d8c3102820fbdae780b68f37ad4a3dcdfd8609042203a,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1475.2741681865632,1475,exact,canonical-source,profile_4e148d1e9dfcbfb1ad710c65,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,livebench,LiveBench,general,dfe97a1bba4c43d4d4f0a7a73d0b448597144234689d67d68de7ce6fea260bbc,f11835e971a30ff05ba2d88cedc6af94873ad0ae2d0bb16e5755a2b6f3b68288,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,77.97244047619049,77.97,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,9d376565bf5dc2ead2e8e2cea56e212ad5aae659f23b764817979b3cf6bfd484,96635d1ad128192a35459af08b4507f194c2981565dc454532c0a175f31f1a50,9b22ead02eb3524c747cd3c6613243d2a3ae31a47ac7ce7e51f01e3edc4c8c38,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,36.24,36.24,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-03-10T21:09:26Z,stale,235,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,simpleqa-verified,SimpleQA Verified,factuality,a3ed2716b4390de63cda63be2c38fffbee5e7ae7d215a8e7b38b827385539fb0,c376dfce5075b65a7c56862e5657658941d22ca53b01ac7bd26dc95ca3f9a478,af3b5b3df99ddad25db0a7afaba9b2d3052c3e051686d2c52d261ca34bed3a0c,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,45.1,45.1,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T19:28:16.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,gpqa-diamond,GPQA Diamond,science,c65424d6da18d166f21fbf969bcebbc0b1ef4c7a950a91705cbdf61c37898110,17623886d22bc314d8b6a7a405988f3a64a2d70acd1a96c23ae773f76c9417a4,21f3b4750ac23c7d045d92073fcbbee018aab65d5e544e8d2991d456a1f06dff,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,91.666,91.67,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,mmlu-pro,MMLU-Pro,knowledge,39a9743fa1de1cc0b14638cd7105eb4b16408e94b473f3016c1717fc47d13f8f,9fa75062ad0ac5474f2da35608077adebb8e65d33c9b5a0698acc3cf7e3a5df1,35280325a2f6f14176f254e8080c9cc9c5655c7411de4981e9b4957ecdb639e1,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,87.482,87.5,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,7938c1e2ae152977a80bba79dbe96665c3ff74898c8819e2b15cc451bb47f622,20beb5ff4191cbcb97149347d63903a5d032ff3f58529d47feaa6f9d930378c4,5700dff2b509c55ed69571444ee543107a87208784ee03e6590b0fbaa8412924,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,95.3,95.3,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-03-06T03:24:03.099Z,stale,240,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,frontiermath,FrontierMath T1–3,mathematics,8d36297c5df51497e950da84dcbc6c7d7f5f476aa64c18471fc238b9ec0e4c40,4b7562dbe17429cb88b8fb007a92504eee1fb88540caa47a7d1fdc5514fb8d62,2a49bf01a775f4a733f04972aced6d25537229525897b14e8d8946d46425c6e7,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,78.59649122807018,78.6,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-06-11T17:55:45.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,arc-agi-2,ARC-AGI-2,abstraction,b25b98eb4fb9bee788833c102caa84bd72aea3b48f9b1fdfe88251a5d8b8b4ec,258c95e02b03547c9dbeeb068b615f2c34778ad62cd471f6546bfc349bd30d0f,b78e285845cfd461df605335673828086cb4d79a24f9cc35618eb474ae942716,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,73.95,74,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,frontiermath-tier-4,FrontierMath T4,mathematics,5eb1b131b31af8a035209d18347603fc859be428b6efad1a1e0e8cf2ee6f8808,70ac6238ebdad9bdbe904cb81679885e8282eac9aaca8ab368d2452156a87488,27b31f681b895e22783a77eed993fda0f1bbc4edf584eaf1a399ee47123a3269,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,49.0,49.0,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-06-11T17:55:45.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,mmmu-pro,MMMU-Pro,multimodal,31ae557ebc10437205e64d620ea4c14d747aa8fdcdffcf7ed16a8102348a872b,cc86630ce9c995e7e1ae8dd91de6bc7c314e06a66e98dd56e4da28ff02bff3ef,d308797ef00d4b8edf39563a9eda3ea3d7a32f03c18b5871fb9242ae903f0285,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,87.514,87.5,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,critpt,CritPt,science,4dd861b811d090bf207add0d6411c7a57da0ad2ed24f5ee9885484892a40600b,91bd50f711ea118509086fa74249cbf7f11b237ff11ca0a4b5ec5bdbf86f86c7,ea2b310e4119a99d31b5c3ff9c76a1ef091cea8acb261867adc06a3b38c981c0,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,23.42857143,23.4,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,ifbench,IFBench,instruction-following,6478a2d63e05d38ba4833bc40b8a15c9e9626e080c7d248bf6c0c1ac7a0c4a68,966dac17bdcc7a606f261481d43ffb77a5aa92151df9604c53f3da35e81463a8,c558a93bada10f3c0600544169f1ced776f37708edc99457d40439adbfefbf2d,release_4adf348ac2f76539ed7ccb10,swallow-evaluation-instruct,ifbench-test-300-prompts-344-instructions,inst_level_strict_acc,percent,true,65.10000000000001,65.1,exact,canonical-source,profile_e80950dca1866f561008c61e,medium,medium,true,fixed_effort,,false,2026-09-30T18:53:19Z,fresh,31,2,B,2,B,0,A+,official-leaderboard,6,official-leaderboard,swallow-llm-leaderboard,Swallow LLM Leaderboard,https://raw.githubusercontent.com/swallow-llm/leaderboard/c58a13a852eba061d639af7caf4ddbdff33b36e3/swallow_leaderboard_post_en.js
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,chess-puzzles,Chess Puzzles,games,312702ff7355d31607aecec115d068948b4ab09dadfcc9e388a6881c1959acec,e49fe61384d80e5fbbfe5c5016ffa6a230989814e4572df9d92e8922814c6247,fdcdeb0159024a2d19cf9bad29a8cd2bec3c7f9133e1920f1b99851790c43e51,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,48.0,48.0,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-10-13T14:04:04.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,arc-agi-1,ARC-AGI-1,abstraction,c68f4d94abc7c045bb1d39af09c4e4f1e90fab82aa9ea4003411fa9c2d53df60,b89d9ae59f0610a45850cd9e58e5fcce3054b79d96ac1fdf44146bf0eaa365b1,69829db519a75c0e24f5b48887b1bc469c3e35ca16a2b02491018075b34490b7,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,93.67000000000002,93.7,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,swe-bench-verified,SWE-bench Verified,coding-agent,8830c3fdd57c563fbe114e3cea9f37fe4577063ca7e93d0e30c3363ef66e3dc5,b31047986d70d360bbc9fe5ece94cb80eb99eab41b400156db5db4dc608c3ee5,0c87e6ff55e71af72ec817666cff3bb6c8377e7522c68ec7eee3fadfa3d46518,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,78.2,78.2,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,swe-rebench-v2,SWE-rebench v2,coding-agent,7eb56263d626318606faf06a0ed6b3f2a926d26b7699077886d1108a08226e9f,1fd030adc5fef5cdf1b1088935b77a6b27fdde33f516138b63c8e85df83a81d1,9ba14dbc07b4734d9269ca0619063b4d06c575561ea00192c192a1576772ae1e,release_84d7ad8aa122e93390d30e06,rolling-v2,2026-03-15/2026-05-15,resolved_rate,percent,true,55.957446808510646,55.96,other-supplemental,supplemental-post-release-window,profile_e80950dca1866f561008c61e,medium,medium,true,fixed_effort,,false,2026-05-15T00:00:00Z,fresh,170,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,deepswe,DeepSWE,coding-agent,65041becdd328b9694ced5596e82faa38d8e071ea75c1559b7a4c40f589b0ddb,74b87fbe5cb9578320ea8f96d94326d160e719665b2a6ddbc54ac844344a2018,a7d2915ef00e671de2b4cb127c262c7196f5469dafa2a20cfc59b38fd001234a,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,51.76991150442478,51.8,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,gso,GSO,coding-agent,141b28111fa1ac29cabe22bed3b30b29cb0198c91fa4f0e28f8a31f12cf07b3f,ec8f7d01cfdf1a0159da9b701935a4af520c490da4cad67fbca5ae4b93d57e10,cfa903819abdeebe393c5cac5f37e9e25b1d0c153e5cd7d29318ad5ecec2156f,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,31.370000000000005,31.4,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-03-10,stale,236,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,livecodebench,LiveCodeBench,coding,3cd553504faa2c8760241526a6479ace02b70772877c83ab2c941674a6b14fe9,ae32c38a5ac096d8a45260b5cabb8c7da89aa9962bf3bfef8735945cf9f209de,04110d937c9fa5d2e2c62e244586e564e413eb8467fbd1c0ff7d0b2944cb4e13,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,84.141,84.1,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,scicode,SciCode,coding,17a5fc4b5490294cce6d99ba7d2429d6fb4723bb46ab420d67a0de9315c9d57b,9befc756c6b459fe8a838ebbd39ff1bce5edf7b9faa4b202c77ee787539b0d48,309df13fb2e5d9a3e71e2cc00cbbef6270e23dd5d77129313564b2d1f25d0f62,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,56.5972222222222,56.6,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,ale-bench,ALE-Bench,coding,c6c6c83670c59671c77bb3b9e7c782eff86df826087d45212f20cf0c62a6c55b,357b5db906eac75c443a5ab4e92bb4c6a9d3d13810546b538158f7e82b172afb,a8155511d9bcff60b3d613babff808e217181f1c4acc9dc6eaad3cb20566ee0f,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1607.0,1607.00,exact,canonical-source,profile_4e148d1e9dfcbfb1ad710c65,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,vibe-code-bench,Vibe Code Bench,coding-agent,edf58376f407c614eac11ce231fa3a70ba32eff207f943b0a161d1d21e3a48ca,c2a1eee4d0c931e57d488726ca52e4ee436b65dfe111018839641785ae9b0a68,8bce169aec517da4f954b54edb939251ecd1bd77886b039de2aad04ec2a008db,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,67.421,67.4,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,webdev-arena,Text Arena · Coding,coding,6378a85ebd342fc603d10ef94c22e5162d2be1893bcf502adb15324ecdf6e867,1ba73b42bb057359904a21cbd62e45848fd8ec1d7fcc68ce8fbcaa576b765c82,29ef6a59018d3edb57c1650e74cbf849ab7ab12c814c4fc0ae4fecc2c857e8ec,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1401.0886143328514,1401,exact,canonical-source,profile_e5f556a0e36183d119678e30,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,mirrorcode,MirrorCode,coding-agent,757792585463f1d9993401d5de37afc923b1e036af7fcdefaf5185088eebc2a6,96b9065c218a2cd958159092001e62224bd61228b07c8409ce54f775dfc39f8a,a501a42ac912129da6c4f1f0956794f1706aaccd3f4734f82fa0024d9ba56b28,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,15.555555555555555,15.6,exact,canonical-source,profile_4e148d1e9dfcbfb1ad710c65,high,high,true,fixed_effort,,false,2026-08-10T16:12:45.299Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,programbench,ProgramBench,coding-agent,45c6bf5477c8c115f35b9cc2fb49adb9fda1fe0f3a7c42351eee62c5429e4fe2,94766ac68b0fa5e2528e7690e6e11d00880bdd7842ca1077ba019fb6f26cb10b,c6ed70bb0950e65097208c035c40c67cc32aa057c56675a727466025adfb2a75,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,mcp-atlas,MCP-Atlas · Scale,tool-use,32df3f8b8cf399c3cf9677f4f95467fbb2a4baff243fbca554c276ebb0c28942,8ce90e1d16ac5d27b64774bf86b70ef26bc28a3466da116a27820936e8262ce7,bae939f021d017dca5f31cd70524b3a062cfb2cbac87240592a93f3c6b0e38d1,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,70.6,70.6,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2025-11-30T18:41:10Z,stale,335,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,apex-agents,APEX-Agents,tool-use,7ea17d74c38f3588a534ef503b36d32d40f6620a73cef04cf43c2a83a95df11a,2efcf64f55b056ddea22b764b71d532d6e40c465c17667d10c8250d6cb21fe63,0797295966e38c58ce88640837d3115765dabd5b52a0e1dc75622c40ce4b7c40,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,52.400000000000006,52.4,exact,canonical-source,profile_dcc0ba793d70c2baa40389f5,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,vending-bench-2,Vending-Bench 2,multi-turn-agent,29927af4509f931fa6fcf3538d7d2c4abdd9739e49f8eaad2303920238137617,fe74521c9674bbf0e26a595ecd51ce734ede4820426661d68a4c60e999e9319b,262915de42a64a04effa6b5004cbc4c44985ca1f5c71ead94e759a23fff7f15a,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,6144.176,6144.18,exact,canonical-source,profile_cf26814a19699b9dfdf18c84,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,42d3de98afa534aebcbae0f0c57030fa9233b3a128f0c38ba0178d5aa1fcd46f,0853c11353a59527d185ba7d2331720a4241bcaa0c5ed816d0bd6c32849bc0fc,4903a568fdc1a3224f403725fe11a5a9fef3a097f78c9b5f969dc67adf188102,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1255.65,1256,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,1628f893717bb52c2f06eb359ba9fe5678ab68c362a3120d471484efff3a13e7,6b7528ccbe40b45ee7f9ec205979d824c64c678aec94c9ebec1131c19f6c715d,1f2c749156758dd2cdd932f2c0fcb8e343f009eac9e98fe705311767aaa43ca8,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,39.43,39.4,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-05-06,fresh,179,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
openai/gpt-5-4,GPT 5.4,OpenAI,closed,25,weirdml,WeirdML,coding,228f85e5f977a5e38c23a6b2ca51e78bc9efee9a4fae1cc9a24e33a314060cbf,6f930aa2170fcbcc26cab0538705f34540a5ce27dc79924aeb313aef62795333,a2419b920174c9dae8ef843b6ee83c439c8799f276a1531e3f5f834e17649ceb,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,77.7,77.7,exact,canonical-source,profile_d50b48d5936735014a249939,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,epoch-eci,Epoch Capabilities Index,capability-index,8b654dfa0bb6b02300ee2769389134483d8698f1441d61620c2bcad2b00b6652,c92980e9047da669f6f93ad6e83308fd05af1d625df5cf60b0da786a94c07cb5,aa701862fb9aeaa7d5cf09aae504091a34c1a944abcae36ef3ba1fc0034bbb88,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,157.03,157.03,exact,canonical-source,profile_6600b24d6b7955f517414625,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,arena-text,Arena Text,preference,e0455f8bed986998de983124434dd9d8d97597f986c09becb466fd36c1c4d83a,4289978b73894ce9c0735c92185cc8c49d87c11b2e4888101e4abbd87597fc7d,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1461.7446227242526,1462,exact,canonical-source,profile_37845fc871b455ab26fa8ccb,high,high,true,fixed_effort,,true,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,livebench,LiveBench,general,f9ad0522dc0663f9930e2a58cb0ab6f4353c860095f6533d9d6d9be67e69f6d2,4d247784bec449462655bb882bf1bed4a6725f5e1aa8e42f50d5c238105a3e88,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,76.03971428571428,76.04,exact,canonical-source,profile_24d6255a6b90676ceab86615,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,df52f43c4637cc137c0de302e32f045a622d5946f587cc4a36535ba21d3312d2,2605369b70f9323d0ce43d5ab7823a5b1c278b4b99eab65a19ceb25a78532f61,f3369573ec5fa708c2fb1887738ca58a6da3b6f6fdd4087151c44904e7e0c723,release_c1afb6c54b8ceec8bc1f121c,HLE original full multimodal,full-multimodal-2500-no-tools,accuracy,percent,true,43.2,43.2,other-supplemental,supplemental-source,profile_4c1dd3ba4a09a30313d8edfd,reported,unknown,true,adaptive,,false,2026-06-30,fresh,124,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,anthropic-sonnet-5-hle,Anthropic · Claude Sonnet 5 System Card,https://www-cdn.anthropic.com/283ef97c476cf442c91d9a37d5b214242a55bb92/Claude%20Sonnet%205%20System%20Card.pdf
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,007bdf4d2939f0766fbe987f8e5059ab32c803fb8c85985bbb61651c6933c3ce,18cd5c41199b06e0b246f42608f3332251f8be263c9be870c0d540ece109085f,d11b49e4e05aeabf126d705998497fd9f0de58914840fc8ef821a3ad6cc23b9d,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,41.28823,41.3,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,simpleqa-verified,SimpleQA Verified,factuality,76e1ee39ac80f3c5f10b5705f85254a989160adcee403938406f79e31e120df4,811f5ca7cec66fb7145925317e829aeeadae09ac694abf97c52a5fe69db1fe83,5861f9a63a01aed2d1972491628359bb645b56f3336acfb048a5a99449787c3d,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,33.7,33.7,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-08-27T19:27:42.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,gpqa-diamond,GPQA Diamond,science,c3065a56d69cc907970650321eb14f6c88320c504f3a6b3a6b20973c12fadd22,ec505904fe4e029702d28c58a1d02a1e3a6a9eb532ea5258d40f358678892a77,71f9cc6e52bb88501e383d272016d2d5a7291de8546458e998292ce0597ab958,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,88.889,88.89,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,mmlu-pro,MMLU-Pro,knowledge,5ec8ad845678b45569d5eeeb0205a230526645b04888cba1ad3908322014fac5,e0e22ef7e05442d676197111faa296024ed14ec1f7d6630d875e1cfc85eb7316,23d9a7703b4749f46277bb2c68fc25c08c0ae1d6fc5058996339007404083e5c,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,87.545,87.5,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,86ba1e466866af1c590528a651c3566b82619f303b73a7e4d86db657fe288c9d,0df49ba275a724af68f42cdf01d8d94b0aa3f9ac788d24b9acbcc5f115abdbe8,9a0868ac73fde87850c5807de8f84ccbf2a2cc8c57e6b1ff438531fc033606e1,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,80.0,80.0,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-08-06T22:21:48.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,frontiermath,FrontierMath T1–3,mathematics,6d2fe2547482cb0b3e9fb0e8bcf78438d497a8f6ec224a3d786adc72d558f613,a67aa567eace0552194963a23b766931e17da26faf6345c7dc46dda72d918d30,6f9eaaa7835e7dbe8796b8a67e624a1fa6c34db8f6fda1c04fc3d72e22630fae,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,65.6140350877193,65.6,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-06-30T21:54:08.000Z,fresh,123,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,frontiermath-tier-4,FrontierMath T4,mathematics,9f7ff9bb9ebbfee2b987676455cf8bf6bba8e88fc53f24e5620a24d718151b26,74cf3918e83600776859688ce0fb683e65495a968a5609a9b7c0963d8a498a83,dbff2f02936d4e952441a601d60d08dd5c4312aaa3c35b3fd6c65947d0f45232,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,29.268292682926827,29.3,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-06-30T21:54:08.000Z,fresh,123,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,mmmu-pro,MMMU-Pro,multimodal,8a5a40287eac36558ec880c5361e38ddd5c020d144df17d2810d46c1968bc110,bc441b56b0bce2079a632aef7403388bc9f72e5e57d3019dd68c67aad59e4cb8,6d8f926de952f64d2866b17d0be9e79385bbf1a99ae6ab095ff54cb8a0fa51f4,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,83.006,83.0,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,critpt,CritPt,science,cfb63c9805ad5d5e87eb6d09f2c29273676dedcb62678e08b40e2617519b985e,1043a4d5a217fb4b356ab48fbe51fb64670aa5063356de98c8e9263bf669505a,c8a8bb2f7d7f4090d5bef01b45161f888bcf65d20d5e142e5a9fa7f6f8e5e422,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,16.8571428571429,16.9,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,simplebench,SimpleBench,reasoning,5f1c3c4e36e86e5b0e7207ea9c08c12734a6c6d81fa596d41d9833836d9760fa,3c22e5025056f2718e065b6c8495fa42dc9dce5aba920738db31d46d27be17f9,cd468ee7668bb194c17795fc612b46112ede3da7738b362f51a2954fff51d4e2,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,60.6,60.6,exact,canonical-source,profile_7249740ffb460d0140f7c11a,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,chess-puzzles,Chess Puzzles,games,302a36e46b2aa4530b30a25a39114aae1fc7cfbb00f0b9568500e1ac7ef763c4,bf8f4594923317c12eaeb311424698c495a5481dc42aceab37636fbea6656f4b,5c935fea8a8010f7c182e214e2196f2f47751199bbbe506cac9838bf3134738b,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,45.0,45.0,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-10-13T13:23:26.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,swe-bench-verified,SWE-bench Verified,coding-agent,3c1f07d258040c7e97fc3de22b06fe1b1eae8e9847a718b3352efaf0256148ff,86aa646db03b80057145ba6e5d6313848e82276831ed960d42edb05f69a318ac,4c397fe6e38b8ac4551eb446d2ee567de90d131360f871eb741c5207d829ef8f,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,79.6,79.6,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-08-28T01:30:56Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,swe-rebench-v2,SWE-rebench v2,coding-agent,1fb2caa7d4decc0d4cc7ea4ffbf25902cd01e205d85eeb456548710b3c69fe2b,64cef58baa68cbe36a9f59574a3c5acd45ef7c5539dd8f08c1e8cf2ea31dd092,94196f1040cab460930695009cb99a0ad38c89a5898c53c3815a04f1b2eea6fa,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,56.75675675675676,56.76,contamination-warning,contamination-risk,profile_37845fc871b455ab26fa8ccb,high,high,true,fixed_effort,,true,2026-07-01T00:00:00Z,fresh,123,0,A+,0,A+,7,C,potential-contamination,2,potential-contamination,swe-rebench,SWE-rebench,https://swe-rebench.com/
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,deepswe,DeepSWE,coding-agent,e110c10f9a8ac05f173262f4589ef8f3131f54d3ea01f9fabc58e424bbeb3400,84c961b29c5c2ac49ea59f3160fe1a15043f1761b204d4a151e8b6fb1a7da2c3,c661237bdafd6d2237be5b74378ff13ee23b9c201c5eec402ab9f08e20a77824,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,53.84615384615385,53.8,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,21ae4e5892a9592442f335c9f8c55ea14406a0e1d77d0294075fba96a0b0ba9b,6618ef96c7be46085afcbc39a382d9d739a6c29443d05b4de45512b6f5c3bb1c,1e4e7ed13670c112bc084e16355db14f56edc12754a029e7282b5a2a2286fb61,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,93.15,93.2,exact,canonical-source,profile_24d6255a6b90676ceab86615,xhigh,xhigh,true,fixed_effort,,false,2026-10-07T01:44:58Z,fresh,25,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,cursorbench,CursorBench,coding-agent,4c33abec313b5204c5eacaccbc0c7b9fb79e3b0039579e3f3d4038db1aad5ee8,59a1f810d36314c4c5b16827ba5b3bee6c7645b6596e37b51dfaa826288a0a95,2e625518e528044afbb746d49b116983badaf752a1427c48d7f9a1ab0857d19e,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,34.1,34.1,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,gso,GSO,coding-agent,267d2717f73c3d4a367b20b9c837c9043072cf1e791662b4aa65c2e40e26b4c6,9c9fdcd25075adbebf3c4b73b4254aab4f8c7f892b7e7aabaf5be552bfe1d438,18b94ab8f540d424f8ed062411389259b95dbf9a9d86613682c99b284484d325,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,37.25,37.2,exact,canonical-source,profile_cf91388c08be94a36c0d7705,reported,unknown,,unknown,,false,2026-07-12,fresh,112,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,frontiercode,FrontierCode 1.1 · Main,coding-agent,cd4d4a4448406f46fb0106bbe9f779153685a7cc5e3b75238e48436b6f0a7c37,7987cdb551a6a637701e730febb2a51982757a0de97ebe61eb7952dd36885e20,f4bcc4c61f272444f58d4638e48d10753414346ab0260cfc3f2575c95d1ca708,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,42.730000000000004,42.7,exact,canonical-source,profile_24d6255a6b90676ceab86615,xhigh,xhigh,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,ioi,IOI,coding,2024034771784537095bfb787fd4d78d5c82fff4962c418007152c59cade5d04,c71342c466083035df905909d7f09a161ed1baf7470ca234d3f487de3c65563b,7b7877c6cd8d806fd5e297f748adf27c6906c2238b73c7eb85666ce58a3154a5,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,45.0,45.0,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,livecodebench,LiveCodeBench,coding,f5022005af7a25a16e2e55a791a1de422531186d70c3376b37ae47d0374a8f14,d6e608d68c588d5e2b4a6094cde74780ff2c4455486a45af862a041a1fda99a7,915d56928cd6c0fb4eeec545994c0ce5e2153b3ea761cfefebecebd425bbe34e,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,82.429,82.4,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,scicode,SciCode,coding,3b55b7cfbd88ec46d5f0012963c1d9e06c36625d6d1aba610a7a2ebf17ff5997,9a772716f3d05c739eb3c7788ac2d0395336ac2af3762e68eeb95a42f4a13ed7,fcb876f2caeaf3498f1d2ff110b7eb80fdf6e8769742151b4e04f79acc0a5133,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,53.587962962963,53.6,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,ale-bench,ALE-Bench,coding,34dfbd3face68fc581570ed7cdffbc53b00a1c961ac47faec3ff16c62843e7b8,ae9794c6f49807909d054b81b884906e4c5cf0afad94b331e7f5b178da00ec4b,7ddf2d974e629787f018a6349f86d2a4ff76f2fb2671251babcf7840475de123,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1463.12,1463.12,exact,canonical-source,profile_37845fc871b455ab26fa8ccb,high,high,true,fixed_effort,,true,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,vibe-code-bench,Vibe Code Bench,coding-agent,ede57d594c814ad24f0001c66d56f7b588c883e5e00a96c8ebc4e61551c6e84d,3bb6a1766847a8e7af1210d390444488899a035000f45342fbf6be3be080e27a,2e89812b273d4614f96bf5abeb68e542a312b55a3b6ba6cc17e27df21a29cfc9,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,81.325,81.3,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,webdev-arena,Text Arena · Coding,coding,0e50effb27440a4ae349b281216e472d854148007f71ac37cf90a47b90f02e15,b00fb3e4dd43b54b5775d9d4b4c759bb68457290f723c2e920a83698798be0e5,24141190a703b11cb1716a50a817cf0992e5b12d13586bac598d26a05f8d5648,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1540.090043634944,1540,exact,canonical-source,profile_37845fc871b455ab26fa8ccb,high,high,true,fixed_effort,,true,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,programbench,ProgramBench,coding-agent,60c450e2dca87353afe3c179b2b066270df8d7ef2ff2000bd076bac4d7177f68,6304cb1a87460ff4f9e809cf004a30c919f97639755d7295d2ebf76f496d91cd,0a70f05d2d4764e578d6f76c9f60026ad2300f607aee73b45ad5b78713fa1eb6,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,0e03b4c7fb76d9aef1645c5c094af0b922036eeaf331a499ee00a79bf56d1365,ef2f154f1a1926a764dcf2f6fefe9d3f7abf12c2af0a12dc742cbdca8be01ecf,1fe860f5c450e0296f4b8821bb3841d06a17dd27afdb4a370e28be5b2c8268dd,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,74.532,74.5,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,c50329c6d2a4f67bc984dd0d1958de45f909657c2d8785efe5664f1a8084cfc8,dbb9de53b9aa49d67da22513e8893544dd624fd3f153776fde05ef403a672a18,cee89bbb1dfffd81c2e6537b0dacfc018e4103f1eb0d457fddcd46530181641b,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,10.606,10.6,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,sre-bench,SRE Bench,terminal-agent,ca8a76a76e4d9354c784f6d11e0f9809b6cd13c630a00a709936b62a1d2cdb1b,7779cd38b95878909086d034344a1173e2dfa83ade7f16c5326ffaac3a4bbaed,ff18c8034eb68af61c56b72d375035faa786f4ec54ed8d6c54a68bb7f5cd4f5c,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,0.382,0.4,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,d8058e3cc20ea614cc072922362db1169246bf0980fbeb69048ea35a6916bfee,a877fd190cf44fefa7c7a4673c2941bc5f813389c9f04a2c4dba406fb952cfac,c0f6bd8ea4403758b061764ef6a9fc9c9d6ebf7a4aa362b3eff4ce5cc27e35da,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,37.319588,37.3,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,apex-agents,APEX-Agents,tool-use,362eee998580fc803ac2397e73f1e74ac358617edb2ecc83d909107f6dad0e73,0e85356d3735f0b8842244161ecdf675c94bdee3c95aa8f3c156b1dce4527baf,dcb2f17904056ad343ca011fe417e55a50597ffabe534a3f0ab7ede9dbe3706a,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,54.50000000000001,54.5,exact,canonical-source,profile_f8d65b9fce47c16194d9f6c1,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,arena-agent,LMArena Agent Arena,multi-turn-agent,c9c4994e60f7ef26ef1157904faafc060a334fe539daf732956d883f0d8b5313,a7a58b18604a7da027f2de847071b007e0bbc802473bfea273d4e466dd43f1e6,9cf8d3866831a13891eff4fae17006c8665285fbe0e1a897654e40b4a30d6a68,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.02940963023579071,0.0294,exact,source-specific,profile_37845fc871b455ab26fa8ccb,high,high,true,fixed_effort,,true,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,vending-bench-2,Vending-Bench 2,multi-turn-agent,b050fecdf81abbcd3d7b33a37acac0902df65e7d788a0b7f6eae541516aaf93c,bd2c212d4e07365f21238f4043c6c4b9c30f2adfe88cc763fd571df295da0679,7145326ec3a37c3c686b6b29841d280a2481e6f2ccf82d77a0769f9153e1b663,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,6377.701666666665,6377.70,exact,canonical-source,profile_145e1c54c865c8920fcac152,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,5006c5024a3586951731bd45e40fe61e7a948cf179691ccd98bfbd442138ac99,92f7b1d2b10eadf5b16037c06f570afb8b656a7683cfaef564e5b843954ed097,7a96862218c72c5a19a246a11dcf15546a7172bc0b53101874fbe0436dc40a61,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1470.38,1470,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,terminal-bench-3,Terminal-Bench 3.0,terminal-agent,5fc4d38e5cbf9f6e67dd7365b8c4bcfda39b00dc2fcde76413c9ee60d8ebd0ad,4448b4dcbe691b01f052cebcc4c29723835582278a9910e49bb06f3a8ec555c6,25f28bd29e55217425ef1850eacad7c52d96511c3d9abb79c78a195ddc7fad33,release_ceb36f25882d8cea2c9d084a,3.0 · b936386b-4afd-4ad7-aa7f-6b52bcab9853,official-74-task-leaderboard,accuracy,percent,true,14.59,14.59,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-08-12T21:39:34.661007+00:00,fresh,80,0,A+,0,A+,0,A+,official-leaderboard,4,official-leaderboard,terminal-bench-3,Terminal-Bench 3.0,https://ofhuhcpkvzjlejydnvyd.supabase.co/functions/v1/leaderboard-read
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,f22344345706e268414163c9aa85edd583d49d8263e806e54a26cbd3be179e6f,cb1dc2002f7b13c9146231fc620945630cc5ff27359d6e04360c4e8fd2744479,bb54ed2d9d35c00dc023cf0200fdc4b97893439ac73542911a208892dd6db315,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,5.714,5.7,exact,canonical-source,profile_289738e1411638bab936a9b2,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
anthropic/claude-sonnet-5,Claude Sonnet 5,Anthropic,closed,26,weirdml,WeirdML,coding,7455899ef5b37b79c2df4db4e5f0b29728c6ca84ec39208e662e308b1eae9c55,107c709304daa8d7b4a73ae19cd6c6d6c26328c26c8a78665b38619828d672f5,19f25720664102c965f733bca8baeea6b5ec87314087432ac558fde76487ce3e,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,68.78,68.8,exact,canonical-source,profile_37845fc871b455ab26fa8ccb,high,high,true,fixed_effort,,true,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
xai/grok-4-6,Grok 4.6,xAI,closed,27,epoch-eci,Epoch Capabilities Index,capability-index,d506c0e27fad55f34c1db82d7e77da8042226188c7296a35a7b03ab835c93cec,ce0c4cbf10c3ae820e103b9182cc51521c9fab4aaea0478ddfbd61e6647cc28e,149f217f6efa014b48eaf55dcdebe61f968caff895e77b0fe2ac433823de791a,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,156.91,156.91,exact,canonical-source,profile_2c4fefa8eb830ac7d5d0c014,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
xai/grok-4-6,Grok 4.6,xAI,closed,27,arena-text,Arena Text,preference,e1939a07f59946d63ec5082dbd5dc02fff579ac2fdcbf67579fe913a17907bab,3ecbeb7d416014624cabe9c71c58b1c4b38370cca0525874beeb63c17603585c,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1455.4817158656808,1455,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
xai/grok-4-6,Grok 4.6,xAI,closed,27,livebench,LiveBench,general,94040cccfcf79bfe932eec9d2ab092a0e742fc524b4b319893ffbc4a8f79354e,b6eea57dd7fef315547e9e206f1d162574b824e68c53c4d57aa8e694d4bc2bba,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,78.042,78.04,exact,canonical-source,profile_16f1405d00db208b1c4a51a1,reported,unknown,,unknown,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
xai/grok-4-6,Grok 4.6,xAI,closed,27,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,2e257f7aa88f6b5e6d16c09c3e676713ab6bbce5e37ebaf0783365e4e2b7f195,c90adc67a939d048c185dd19ba8c3a7c17678fb4c75f67e58339d39270de659b,cfdbd8716d06d4eedce197dbfe61da5d07716d5493fde410304ed39f2f4a3e62,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,44.068582,44.1,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xai/grok-4-6,Grok 4.6,xAI,closed,27,simpleqa-verified,SimpleQA Verified,factuality,75ef1504ea6d448658071128034b2727ff70fbf52defc8336d7813fec909dbb8,f84084e21d3b0c30e46cf9b41e5b4bbef77cdf3f77d5cee052c13f7c1791eedc,da916f7fcb980502d9849f78cb0d5296fe195cadca371274fe57d0a51007ab2b,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,48.9,48.9,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T19:29:19.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
xai/grok-4-6,Grok 4.6,xAI,closed,27,gpqa-diamond,GPQA Diamond,science,b258d959bf041dec098133264f0b62c0b2b455a0ba7d66bccbe1d9bdbf7d3bfd,8dffaba90de43e133da96228ab296bc0da0c67072af3227b0da56729fe71de7f,c13fcde4446d6a6ff0bf8d8d0146b862a68bd29478d6cec5cb871fa8a72c9600,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,94.697,94.70,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
xai/grok-4-6,Grok 4.6,xAI,closed,27,mmlu-pro,MMLU-Pro,knowledge,99b662c0dda473d55aed96672238c814e9cc9a9006b667f001bf991599c704be,e20b1451a6bedcf0b124868488a586f434575458ab0e81adfbe4be2b93748836,c238afa94e515e843c09035f27351d1e0b396c255da87edf0172065d11d9dc3e,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,89.4,89.4,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
xai/grok-4-6,Grok 4.6,xAI,closed,27,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,d837ba58f98f3965397b3ec568984f7eac97b316d24d082428ea3547d52ecc41,73e1a2f5c882e984c727799374e51e5de3210b8c14ef884bfca233af43e13e33,c93bc277c165a94c5c0d5cc658ffaf45c61cac8abd5fae1685dd16212920b64e,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,99.16666666666667,99.2,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-08-14T13:47:42.000Z,fresh,78,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
xai/grok-4-6,Grok 4.6,xAI,closed,27,frontiermath,FrontierMath T1–3,mathematics,c457fc2646fffec8cd18e0fd18c22d5c90e2ed52e8c39370ef7a999bbb9a66da,e7dd26eccd45645f2383cc60eb1a1b3c6e0381883d43062199c83a812320ca2f,1d70122469f92bc372a5aa8789c446f32984b20e1520f06417885ade1646b058,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,65.96491228070175,66.0,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-08-14T15:46:50.000Z,fresh,78,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
xai/grok-4-6,Grok 4.6,xAI,closed,27,arc-agi-2,ARC-AGI-2,abstraction,96b4262b3ca65d762ddd216b89564c618e2fceff025cdee3d84d34fc0c9234c8,5ef851b8417d4a6c2a65534464f51f84a793cfdea9adaff0b49f718fa6f3c7d5,f2deb1a5fb373735d8e5af609ba2b55d428ea3ed188e7920f86f3a3b19d1be5c,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,67.08333333333333,67,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
xai/grok-4-6,Grok 4.6,xAI,closed,27,frontiermath-tier-4,FrontierMath T4,mathematics,414dc1a81efcce65fb7bed778f0d6696e9bcd5718483b5da647bfc036fb94ba2,8dbaed9c3766375f3bc8e7f695215ce7c62b0f8c31f8d0e145dd6d4f7f4c23c9,6ed3f071c5e7750e1f2076ddd705f98a41b812db3c402e1f7fec0f3ba3bb28e3,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,31.70731707317073,31.7,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-08-14T15:46:50.000Z,fresh,78,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
xai/grok-4-6,Grok 4.6,xAI,closed,27,critpt,CritPt,science,de465e8ba9b2aec95be1ee536932aeb9fb7bfe545e7c6f78e0b16f13d01f1d83,2d10be2e0e8c6bc67d3b960994c9716637a9a43f62bf5dd4685429451516023e,02e1e40fd693a38356a09be1fdcc0d17fec0ae803a060cc84b5343877a99af64,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,19.7142857142857,19.7,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
xai/grok-4-6,Grok 4.6,xAI,closed,27,simplebench,SimpleBench,reasoning,5a16f18bf9de515711e931045d46170eb32b7fe983d9b3e5d65c5a5477f7b624,c9146da17f50da205540536c98c535ad8635576651e75664a1d7c5d40ecf81e1,6333f90af491572b991ae0a443247ffa5468890e9143e9ab7c4cd0c3fff81cfa,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,75.9,75.9,exact,canonical-source,profile_9702996790d011ea744f76f7,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
xai/grok-4-6,Grok 4.6,xAI,closed,27,chess-puzzles,Chess Puzzles,games,a898b1220afaed3c63acb3152cfefb9bc5fb3dbe363d95f4e577add44ac1e69f,7c45647d4f11527c626472ed311e29c2e551a700ad425ef703093cca4de9ee93,e3f7f249f7399c2a8c0d9af26f62a0f7271255c991616b72c65b2265d00e5297,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,33.0,33.0,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-10-13T13:14:03.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
xai/grok-4-6,Grok 4.6,xAI,closed,27,arc-agi-1,ARC-AGI-1,abstraction,d065a43663d65c9945d3840c0940eb3eb550fa560ac9f14244c6b07847e4feb4,2fcbef936bfad58617cea550bccb91ca868e005f9c15f6c1f698da177fc8e2ed,67febf66d31d5521c9eb0190150dfdad3a229e9ce81cef5873103d59a6c61678,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,87.0,87.0,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
xai/grok-4-6,Grok 4.6,xAI,closed,27,swe-bench-verified,SWE-bench Verified,coding-agent,8236fbdfbf810d9832dce5956d7a2bd0380b7f6be56743ba7636d8d9842195c0,e365639172ce2cd9646ecaf47f94cb660642dad1333e85aa81bc7fde09dc7b7a,f81711268b3c272ae18ed117647dd4ef08a0fcb2f8a8f6b28176dec36e465d02,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,95.6,95.6,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
xai/grok-4-6,Grok 4.6,xAI,closed,27,deepswe,DeepSWE,coding-agent,c7041244874e56c537e91fe46a135aa4d9048b753fbbdfaceb0d592e00e46787,8c32674f624aae890cd7972acb429e7445b7f4811a32e542dcd723e9e6ffa727,951d29baf7cfd97c4c65aa4eddbb9245d35b13cfb7b196883aa8dc2e169aa296,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,66.74057649667405,66.7,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
xai/grok-4-6,Grok 4.6,xAI,closed,27,cursorbench,CursorBench,coding-agent,af4f8da2100ab931efcb850beebef9477058f56b4ede4353aa0ff89677b6bbef,e62c80f6a0331e8b59a20d71dd74f237c328d2808485389c127c535a2abe5c76,e3754c0d659acd22edab1be7e7103caee011039ded39b4c5528aaf07dba237d9,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,41.4,41.4,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
xai/grok-4-6,Grok 4.6,xAI,closed,27,frontiercode,FrontierCode 1.1 · Main,coding-agent,095e91129eeb6dd5ddf8efa114df0a5aa297ab31feb02d44f7d35730dc46d820,8f78ff3ec5c52a4a78ff03fea09d0c3b761bdfcd0c7ac0c974cf9b72f3c30851,5b70a8512e640283f3787f677a72cf593d4e73fc125f674ceb0c28906a2bdfeb,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,48.010000000000005,48.0,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
xai/grok-4-6,Grok 4.6,xAI,closed,27,ioi,IOI,coding,e77157d94d32664caa3d13599edc52a3d73f0e60d21bd0e58d9f166c889bfe52,21ba530bb0cff724ad20af90f125f29bd9c6f8197dc66527eafc717e656ba93d,db9af69fce7968024fa9e31b3d72c863c46c068cbc1d68aac93e81e5d97c7700,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,47.611,47.6,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-09-14T23:34:42Z,fresh,47,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
xai/grok-4-6,Grok 4.6,xAI,closed,27,livecodebench,LiveCodeBench,coding,d5402037f631f9ed4bf7012905e2dd6a89b8623bbba26af1c548af82cb330ed3,f2ab4b636c9bee47f3d7a767f496384b101bd27213bd3f6f3e3265729985c5b9,0a315b318696d5c83e1a5b03c9fa7c64aaecbf69e5882ca81a74429f866386f0,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,88.224,88.2,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
xai/grok-4-6,Grok 4.6,xAI,closed,27,scicode,SciCode,coding,3eb8a6649fd58f62cc587179502f697d8807833b29d1f2d8660022008540b04b,dae123af0fe09d9b0d5a26689e7eed800e1a2536f58a86ab5ac1f73a3ff5dc99,78ec003137b0540953266132999d87a5b81cffe8adff60792b3d1571152750ca,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,53.0092592592593,53.0,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
xai/grok-4-6,Grok 4.6,xAI,closed,27,ale-bench,ALE-Bench,coding,e760db9617923047b9aa7472bf6e51b435737f43473ef20d2e43dacdc46d8a5d,d2e6c16631cbe712cccfd2e84b4a7deba258c4f82820d1a8f342af234dd66a2d,c30da6b2ea8861545a89073da5255befc4b15d43db77b3f5292f4dcc43404be7,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1508.08,1508.08,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
xai/grok-4-6,Grok 4.6,xAI,closed,27,vibe-code-bench,Vibe Code Bench,coding-agent,7db9c879364828e26f3e7543a218fd9c2064b4aca4ccc2874603cee569f0dd60,813870c0494145ce46e9b34ef756948e0c0d9bb44c2f7525459c6686d0a2c12e,38d83417ee8b353264646e28c5708757ab3cc92ceee3eddafdd46f26512113b4,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,76.239,76.2,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
xai/grok-4-6,Grok 4.6,xAI,closed,27,webdev-arena,Text Arena · Coding,coding,d9ec62881b507eda71a8e5ff9cc5fa2d049d663a0cba73fefb5dcf3d7e4f7e93,32d6a38a21887795dc4cce8e1074e3cdedf70bd01a5f2fed0360b390caeb5bed,3c439afe3cd7d7ae939af8ab5b79d883c1c1b10904ee138f6ee235176521c39e,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1616.3867226095672,1616,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
xai/grok-4-6,Grok 4.6,xAI,closed,27,mirrorcode,MirrorCode,coding-agent,a59e2d561518fe62afc7fe8798ecb12379339937b8d5272ed1d4faab2e6eff4e,c6117af0569720118381afe03ef0b8e31af635706166995716bd28b647a0f16e,54d4c78e254afafd6c54d3ab82a18e60308491b654ad93ca9f9d0e896cc59b43,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,41.11111111111111,41.1,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T16:48:25.000Z,fresh,10,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
xai/grok-4-6,Grok 4.6,xAI,closed,27,programbench,ProgramBench,coding-agent,e235f7cbcd9972dadd6830ed750c4bad9ae452b5cfff3a0b075b4e7ff5e71295,8debe7667f85c956498fd7cca528c4785f4852aee4ace97abf5fcc36ede060c6,8d9925817fafce60c704686d6bd4d4c6f16de1dc898c939ceb740aff75c03afd,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,1.0,1.0,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
xai/grok-4-6,Grok 4.6,xAI,closed,27,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,1dc3ee798592e803269ec80f39cee2ffcf61c3f005b63546849dbc8443350f22,57cd945b688d5dad9746d0bf643cb6603da4229a20773b1042dd0fd37ee90027,24d04cfd45fe816969d620f9b1a0e9abae525a985d6b3a68b2e43580a9a020aa,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,78.277,78.3,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
xai/grok-4-6,Grok 4.6,xAI,closed,27,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,bd5bd3dcde731797e93deed89e5ef04cfc490598189d3787ef1c198c2c244acd,da74a87cf0df39fe3a06c3f3d3a9ec970f3083098e4852a4bae5f83354a3c1b6,1e5e078634102ec500327d1141cc3106a52938d355d6995fdf8e0ee54a7e3115,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,17.172,17.2,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
xai/grok-4-6,Grok 4.6,xAI,closed,27,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,0c2f6e69a37cb9c310666486c35904f4003687d28c878cfeadf259e66a27f59b,2ed5bdfa5ee7fdd13b36a56dd668e47205f0eb44844e4249353ed6c826d83403,849330db086932ad1604dfa534e630eecb6bd61c5a7d8bcded83bb69a578f5f9,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,43.298969,43.3,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xai/grok-4-6,Grok 4.6,xAI,closed,27,apex-agents,APEX-Agents,tool-use,4bf6d535b7248e39a3f8f840e30dc1bc7481065533a09206cd085a5c8bfd7aa4,90413110da253b13cdd69eec5ee3e362b034051470b13964a3099465b3712fda,7de066374ca85de128d967ed102edf469b77fbbd2bc137abd8ff65b2409bf0f4,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,65.3,65.3,exact,canonical-source,profile_42c5eb67e57e5181f3df3cfe,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
xai/grok-4-6,Grok 4.6,xAI,closed,27,arena-agent,LMArena Agent Arena,multi-turn-agent,d96f06e478678c45f97e0e5f5689a3993357faddba3e230ce5bbe7c2d9c37235,0ed6552370699e5cd2481a9a1fe57baeb6161a882d51f6a5e5788925214684fc,4a79c613fb4674130515f1400b113b92055eaa183769dcbd5c95e124fe88de47,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.004874136743475029,0.0049,exact,source-specific,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
xai/grok-4-6,Grok 4.6,xAI,closed,27,vending-bench-2,Vending-Bench 2,multi-turn-agent,6f763412f1a4480c09ec643a48abcdb609af716176a0e21539be788f91da2049,4f985f16a80fa7feab98175dc7016670de7c45f511bb87377f00e23209d40162,a904ed276dd3469869303926a854441ebe2a2f6ee78e0ccb7665fa02faf182d4,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,9047.032500000001,9047.03,exact,canonical-source,profile_acb04e327788bc5d55cd1774,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
xai/grok-4-6,Grok 4.6,xAI,closed,27,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,8bab972accd291166cf2c627f6d2b9e5edef6154cd35e34fd4ff2188208cd3b6,2ea02a0268777204f10388ec0c9ee45c902f588a95c8c091292470409dee7278,e51220aa02b5a4a0606a516249103fb5168f4b99d9480c17c405094b1aed30be,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1654.15,1654,exact,canonical-source,profile_466c9201a70dbe424a06eaf9,xhigh,xhigh,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
xai/grok-4-6,Grok 4.6,xAI,closed,27,terminal-bench-3,Terminal-Bench 3.0,terminal-agent,ea737046f2ed875df0461c20cf6120ade4fcb00c85ef37916db65f211a2c46af,0d7ac4f6331e114e3c830017f0228b9791c6933cdef6d873db8a4695bd46724a,cf3c5bbd00982510245a76df7e2e0a8eafa12c3cdc98a1b67f186bc10a9160d2,release_ceb36f25882d8cea2c9d084a,3.0 · b936386b-4afd-4ad7-aa7f-6b52bcab9853,official-74-task-leaderboard,accuracy,percent,true,26.49,26.49,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-08-12T22:31:18.697086+00:00,fresh,80,0,A+,0,A+,0,A+,official-leaderboard,4,official-leaderboard,terminal-bench-3,Terminal-Bench 3.0,https://ofhuhcpkvzjlejydnvyd.supabase.co/functions/v1/leaderboard-read
xai/grok-4-6,Grok 4.6,xAI,closed,27,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,7147c7373767915799d8fcfeb54cd707cbc2dde03e9442e50e30464d7f44ad40,bd332329443966e6a810e8fb945051b768ead9d12082d3e02ecfeff77cbd0530,ea7aa94fce34860455d5930f77d7f0779744e5c64f59036b79587b297dec79e4,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,11.429,11.4,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
xai/grok-4-6,Grok 4.6,xAI,closed,27,weirdml,WeirdML,coding,d0af1496283121a4bda9b5fabccbaaddfc1593a45025ddb1a13ea70fee02fb3c,9dd793496fe7011a8c6e46476866e47f600f29791c59ed737043c49d5999a875,97d8f516d27f31969b45f7d72b7d15b3004d5a73f2cbdb2d0d84de6d1740d7eb,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,67.29,67.3,exact,canonical-source,profile_bf39f31a22c70a3b0ed8abd8,high,high,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,epoch-eci,Epoch Capabilities Index,capability-index,72ce137cdd70c6098873dacc4eb585dc531d17c3cc43c11783f10ca75b6075c6,2ee3df58151b83ed3681eb63ec5e1c3904816a9c7d1b7aedc6de10a5f8ca3269,234c9b3e03a9180a88b15cc6cb61881a1465a2879362b31247cca51149788edb,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,156.75,156.75,exact,canonical-source,profile_f89a3ff70a89c2059aea00e4,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,e09886d7e6355da790321c6149d40caaef73ef17a58d9f305f3701df194fba4b,494feb1723c78a055ad16b4b6880740c3d19d7eb9e121d7ca51954a38076cd9e,6ffd600efa000e10de64714a486bee8c481258e5f579139fda551925122bcf6f,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,27.7552,27.8,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,arena-text,Arena Text,preference,bc3b8b1de789210d94546f46918673e69a257f5513f92a29b4fde865c4558c0f,06c0a110759dc893e920afd636e66df20c374e8a5d78ac155a8f759004516dab,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1440.8674264272704,1441,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,livebench,LiveBench,general,8fcedb4d33559e8cc97f3dc7a373238b46b044b6bdb5455b51aa8160fd54453e,ed3738fc571fb52e5bd13f0071816bf3890b8bc8073b4b03d486729f70619117,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,72.02647619047619,72.03,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-09-30T18:53:19Z,fresh,31,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,8c7c223d97cb2b7760a0681e7533d6afb2edbc25ecba5f1aa33121babf5c0274,128403799703a9a8a4b2aaf71b38e246295a98afd8e0d3aadc353ca243564e2f,1f7d8f877d274bff3d1936f26c62e7a0cd51413f01b278407dcbf65da8e1ce74,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,38.507878,38.5,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,simpleqa-verified,SimpleQA Verified,factuality,0f238a1f48e79f466e9cc5f30d6c1f4e78fb659a56809f1f065e4040d74e6443,7b9428dc449b5d11d3ca4d4a1de86184855ac256e07bc69c09f59402cea94fab,e94db4081e6470cebd76cf723afdbee751fd5ab78aa7caf1dce8051428180c7c,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,41.4,41.4,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-09-22T22:14:05.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,gpqa-diamond,GPQA Diamond,science,41ddbc414abecbef3fc362f3d176e4382f3573674e8da2445b465b6200a915f9,6b091a572b5dea41b3780605c126f851a6ed77e7e051789b471043bdc0cce284,4f8894782fe7146bdd6e7823b91e16cebd928ac74a010334a517c63a28e2063d,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,90.46717171717171,90.5,other-supplemental,supplemental-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-09-22T22:14:05.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,df00660f1bedf20388788b9b95d5518e8fe1f37679643c66c44280fc3d45cc98,e985dbde71475e7d462cb8ed985df5ba6c5c3e4296e3d6b3432a9a620559833a,f54e9344c5dcccd9c9099f125e8e3dfabc47aca4dc7b6a2376d820bcd2bc5cde,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,98.88888888888889,98.9,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-09-22T22:14:05.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,frontiermath,FrontierMath T1–3,mathematics,15085814b4109cc7e5040e5998967a7eeda3ff68a773a06a27f870c5f7f2ea2f,e6905cf6cfd7f15411d63affa2df1a6a8a10f0270a1c5b6ab49de46e3adbfb06,afddc02d7d2d96b25d4a0797f0a91eb3d6acc65cb504e2ac84997901fb605d67,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,78.94736842105263,78.9,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-09-22T22:14:05.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,arc-agi-2,ARC-AGI-2,abstraction,795c290b8c25680d491dd81b040900ed7c15ab9a3fd46861360bd519f74ad5bd,45393279f940f58cd0467d4d5f0ad201543cacd0805f4e931f86448659be7dc4,acc7fae2580193dd8d8dda87d8933ecd7f1fc199f9fbe30a6a65459cc2f0fa72,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,59.30555555555556,59,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-09-30T18:53:20Z,fresh,31,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,frontiermath-tier-4,FrontierMath T4,mathematics,dc0f11c97ddb4b659e3f200cc9fc99c8d4aa292be7ff65f699382edc9927c45a,c88c3fc43ac8e0a2a4b8c840fdce5de1c917993c23de81efb377449dfcd7a791,521221b6baae86dd00715ef32c0e7fd47ce2cc18b3abe424e51a2346c132866e,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,56.09756097560976,56.1,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-09-22T22:14:05.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,mmmu-pro,MMMU-Pro,multimodal,c0b8e9a1fcf3e011eed11a2392c3169b3e9ff3d0a5149b62d9bf5bb06de430be,74c30619d54051e1ac077bcf11d23efb9660613b173692bc1efa846f545b0118,06302ee11935871f63c5755b7d04257ccfa327f736e4e7c292bc3bdd296f64e2,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,79.653179,79.7,other-supplemental,supplemental-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-01T09:31:57Z,fresh,30,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,critpt,CritPt,science,af1759b747fb2371a6f14399b2ab9796696775a41b8fd8c583eed901f8a4c247,85d51605446326061aca78dbb6c5a9561b2814c7d17ec6f9f0f9da5a5d23f030,f1290bc68b31cb81971ab6b7a7ff1de1a633d71dd1b7c6e10adaa7be0f0a2d21,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,19.428571428571402,19.4,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,chess-puzzles,Chess Puzzles,games,4a707e2253ba0cd0921194991d1ae7617db6b218e41a06224af9a1965bda6e37,a4de6cc3cfed96284b6eac0f57cf6151f7a6466f5bd7b5f30f7ead1bf11dfadb,e08da095d58a95c9091dee9f5c6a93caa4c6b0f281204720f5e8776ef081e51a,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,36.0,36.0,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-13T13:37:06.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,arc-agi-1,ARC-AGI-1,abstraction,04c524067ab8c581a9c387cb8da0dae9e9dff55387e0b1c908fbf2f886219e31,0b3921c0379dacc3bfb6c7eadeaf40f5a4c5285271464e3daacafef68f71f269,eed432f169eff7b8754f49763493b877a3d28a5bc99b079c4137be046585d784,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,86.66666666666667,86.7,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-01T00:44:57Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,frontiercode,FrontierCode 1.1 · Main,coding-agent,7f7dcdda4f46afdd60236e67714fa925506b5d6e74468757b4013b1bc60421a2,04e2c930c878264e31ca5cbadb889b5d891b4c1382bf6fd6740db8d037341fb8,0a6c6cc0cc8cd63bee36b5a54e5a6696b3e36023dab704c63cf28436f2c91adf,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,42.42,42.4,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-01T00:44:56Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,ioi,IOI,coding,ab33740e10842577d841872d8ee2bef4d381bc885d2c9f835e5957e612d60efb,bf3745f8d0ebf59ec31f40137629b2b8b02f60907c04484040c7e0259617aa09,47fac0a651f82a10bce6311cd89567a2e78c60765370744609e45aa13c63cdda,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,55.556,55.6,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-09-30T18:53:22Z,fresh,31,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,scicode,SciCode,coding,45af8f467debac540282971e104a397698fa736ec9268618deb89ba30bb216a8,2f64b38181de3b7c262a79b0f52b8bd319b004638c1cff4974d4313f100676d8,3fa5d2fd06f1e529029ee43e1ca970898dddf947620e3d97dcf61bae69a4ad2d,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,54.6296296296296,54.6,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,ale-bench,ALE-Bench,coding,3d615a3401f14e269d1943ef450debd7ec5c3385a357fda2efdf3bf409eb3477,bf2bccf600bc93ff68a17d58b384f0bdee4170448f9bbb246885291533c588d9,187f849bb54a75712c9a6838bd1bb2e80773d3ef45aa8518c75a7ed32b8d41bf,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1576.92,1576.92,exact,canonical-source,profile_c3fe93be8d4fa0fd59bc2d19,xhigh,xhigh,true,fixed_effort,,false,2026-10-01T00:44:58Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,vibe-code-bench,Vibe Code Bench,coding-agent,4c1016408b1000896481170b61da3a9d87ca4ad49188be0efa5237e59b9d802c,7d85c63980d57ffe861021cb90be62a164463b38febd2d4e70b405badf8de3c3,6990ae8177f7de7e651ad2f7bb9b3bb7275dab96423fcfb41783b28e9df4887b,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,81.649,81.6,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-09-30T18:53:22Z,fresh,31,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,webdev-arena,Text Arena · Coding,coding,d4643aec594442712a5e2da11346346c9e871776ed7ebe833fdc5a23fac6de43,ebbfb769276be162ac8f0ac3d971b55b4c4d0f4c75e0c35db7226e34ae3e2de1,65bb3c6871b7029c925b6cae53e96692e7ff08278b3f245fe85e4829a4d9c47f,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1579.4041580064886,1579,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,mirrorcode,MirrorCode,coding-agent,8895df96ff2fad48b221e5e48ae33a5ecd45bd555bb215c1892facae2ec795b6,c3c9f23b6c468b57f64377402ae69df6fd8871718c180194678428583641b9c1,1601ace96b6de60c5987767c19084e0d476b0a5ed13701b20147bc0650d63f00,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,26.666666666666668,26.7,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-21T16:48:12.000Z,fresh,10,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,programbench,ProgramBench,coding-agent,544bfc19678add41f7f95b6485fd8b17c31f91ecc334fdf85640ab471b6d568d,ce8036c001cdd89631bf451827f6d9c5bfa208fe473d56d7a392d30b1e757837,49ed034cb414feddfc6387d7b5dd76669b7aa8f94e2c9d0ebbd45d8f6d012c36,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.5,0.5,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,weirdml-v3,WeirdML v3,coding-agent,bbb54fc6e87e82f628a303b748dd03eb6b53eee3c8d6cc2acdedfdfe502c4e6b,8f5651919ff24cce6c7e024ef8cfafcbd3a8badb622318c8991be6afb234f4a7,b8828ae6627c35958f7f5f7a7b4a1d268687438eeb1a8644428696400b62e285,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,7.874124242424243,7.9,exact,canonical-source,profile_c3fe93be8d4fa0fd59bc2d19,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,5e9385784f690a1e7c363decd3e17b20741bfb056dd8a00c59219b180d7fada9,2d19a6065a7dab7175f956c4caa3cd701299ce6b872cc75aef92e9c556bc71e6,b97b7878b69b69f88b14e51a1b50b7590089a145a20f3e54d6a7b75df4842fa6,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,73.034,73.0,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-09-30T18:53:22Z,fresh,31,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,9509276b8ccac1dfc11443e7e542e749daf0d8b2eb307e27b5eda134ce2cb45c,d4f753fcb2320ceda48830d9b4e16e1d67d34d12f1ef8dba418f38e01d98a4b3,34846358d5011649347f498550c1885ea19b167ac24b18f5e10c54350500ad3b,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,13.636,13.6,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,sre-bench,SRE Bench,terminal-agent,b73da222b60d7fa93f4163a36cb6d9db211db8215ff52104bd4134f1155f86e8,ab403096f5baecf83430556ec0a62ebca536e9e5695936f98c31087177a34534,44257bd57b7d53c083cf30fa16edb50764c41fd55c5233eee3a88fd85e1451e5,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,2.672,2.7,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,apex-agents,APEX-Agents,tool-use,a16e2fd04a73bd62231c79b058394a3d875ef21269a2dd48ab4723305be8d4b3,d6b1ce8006a614db7ad817a933f8e6a3cd1810ecf5c890299650fc52f8a22201,209c3d5c6e7827761c8410e15af99e9f7e7e411b29820fbc293b2ac9069cc81f,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,44.3,44.3,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-01T00:44:56Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,arena-agent,LMArena Agent Arena,multi-turn-agent,345692897f8bd0a4beb7f18c816b28b0823531f673aefcf55188087574f64816,fb49145da4ceb168ba89c13d96fc5d5300d600f21889f04ee7540e59c20cab05,a3509388d1bf9f5b3b0b647b1835a5b447a2c0d552cd02567dff967b409c69f5,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.02225749002662884,0.0223,exact,source-specific,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,d4abe2c8d3ebdb6a36b3bf0666bdb526aed4d8a024bdb951f0dc74686787d8fd,ac55262cc6bd2957b4938659af0137f51267d4c2eec5b3ff3679306fcee1bb78,b89e84b47d5fc884417139aca664494be0c6374fd97530e1cbc3ff383f574a23,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1444.46,1444,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-6-luna,GPT 6 Luna,OpenAI,closed,28,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,93b4e062a592a8a60084ddc5d6e27ea7114572efa2e8ed364cdd79000e6f86c4,925ff00c9ac33911632ec3c9706b902477ef67a99244d092c13464158e171f80,5c91c0e98fc01eafd618671ec69f135fe0cae45189ca7eaf35e078e90ba211d4,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,4.286,4.3,exact,canonical-source,profile_c5c378717d51c7077b3462a2,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,epoch-eci,Epoch Capabilities Index,capability-index,cd2c523f677a97c852db9d7faa5614ddcf62b120b8e8812a665d1ba784ec6cf9,17374bb4ec50287ac88cd88e6d677697b5c98672e505e6f0b79afb27960c0d06,f57751b0fc61399bd3c3114de36e13164d73dfd2eb59793f912ea07282dfec2a,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,156.7,156.70,exact,canonical-source,profile_bc89e06c89d745c77565b69e,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,arena-text,Arena Text,preference,5663cba391b4a5fc1d1af87f384ef087ab93b8bdbc7204b81b84b6e616e7e444,9bdc6d0940f209ece7fa9f450d2eb825eaf66d4ae3490d2f727f6d68da712d34,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1501.0497826033693,1501,exact,canonical-source,profile_1c905f9be1ce4940b60cc62b,high,high,true,fixed_effort,,true,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,livebench,LiveBench,general,2af77b2d33cf275249d6161ab0a6acabbed4f4ba677063e38dd780f43163e678,2293a9b14524185d1921c9b46c48427c811977172257ecd0713dc45957db77a4,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,76.52935714285714,76.53,exact,canonical-source,profile_509db2bd77140fb589d8a000,xhigh,xhigh,true,fixed_effort,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,ee8e4576e838fb909aa9017714695d49e5b11de5e10e8fccbe0d466050d0e53f,414da0ddbbc9d75652f11cac5919889f7aefa7890d4db23ca464a035ef224b1c,086db8f76fce15006f0ddfb02f28250a58378faae73ef929e2f0e1ce6ae5ec6f,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,36.2,36.20,exact,canonical-source,profile_706026ac74b842322735028b,reported,unknown,,unknown,,false,2026-04-22T20:13:01Z,stale,192,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,simpleqa-verified,SimpleQA Verified,factuality,b43fd2425a7905a25019179af7fb445aee4acec37ad8798e796242c817130a6a,360a5deedd1ce1302fd5c20afe7337c2631289774a68e33839b3fe66d2c2dfd7,f71750a6cf185b750ae565540c105f363a1e1c1a10d96ca8f6af6b68524e1165,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,51.7,51.7,exact,canonical-source,profile_509db2bd77140fb589d8a000,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T19:27:26.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,gpqa-diamond,GPQA Diamond,science,5c1d601cffddadf02729039030d3718677ebe5915a4b295504f0aaba280ee844,2ed5aa681bdba878f745ed49f7ca58f7cfa898cc520ac219f3a70b75b4c0b98a,728264391c56a8d3926c2278c00792a8b7b88e8f70764caee51e1c0c2ebc63a0,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,90.152,90.15,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,mmlu-pro,MMLU-Pro,knowledge,2dc834162c5935edba463f2224ee000463744cfdd3c6472f633acb97708e17ce,21c46aa232f825c2649b841669ab3b316a39066c0245d1758e57e7c05259fe95,68edbe85ff5ef67d9841aa9ddc87ad34f8bf362405194609df10299abe8521c1,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,89.871,89.9,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,0b410752a4fc70910bbf9366b54de3db5aa100de62ceecb819a53927e5ad518b,229fb1e24a6c8898529f38d2eabf804fe8de7503e148fe55f30e21225daf1bcd,783bb1d83e7d8cebef6c7c4b8de5e536862f4da779b369aefb639a940dfaab18,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,86.66666666666667,86.7,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-08-06T22:19:46.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,frontiermath,FrontierMath T1–3,mathematics,3cca82bfc2e31c1e3ef47646d76ff3fde98c81642483c137d7ecd461d9f772b6,3bc19aade29220eea03c05e75c099cf382eddf2d6ba3c9099fb64c2573bff137,2af8f172defdd8b9838fc070e6004727cd471180369f89d718b207d85964416e,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,70.17543859649122,70.2,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-06-10T21:41:54.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,arc-agi-2,ARC-AGI-2,abstraction,c852770962a8212b71c7e50843eead808829678661fc9b62fe2a4ae127ae286a,e2e1cc58373390f67b0b48c92e53f3c49a3c404dede08737279234ddc4093319,9600efb6cd1778acbe3b03291aa2b77de82c5d3cda92d4d59fffac24b028bba7,release_c66f5b82bd3045b91ad807bc,ARC-AGI-2 current leaderboard,evaluation-pass-at-2,pass_at_2_accuracy,percent,true,75.83,75.8,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-08-28T22:43:25Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-2,Epoch AI mirror · ARC-AGI-2,https://epoch.ai/benchmarks/arc-agi-2
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,frontiermath-tier-4,FrontierMath T4,mathematics,e4209f74b4d1092572b29f991a795e7c6794c7dbd936717393ef3c20968e9bd9,03fb332fd475adc6666362391691695b20bdc6216a4b438d9a8b20286f7ba533,4b2aa72f1aa2a2342d6972ee79194b6b97e00904c2418a36cfcd8df4fd9e170f,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,31.70731707317073,31.7,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-06-10T21:41:54.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,mmmu-pro,MMMU-Pro,multimodal,d4ad3c12a6259df427ba248724c058bd1efcd1766b8aacd0d81d1e4e75c39c68,4bc2e129f74d42aac76cd883ad778dc497617b80e6f378b6a2e657d3f259dc35,b4cef566aa48377ab1b8d58b351f9c4d8b19afbcb7540742ff46243b6eada7f4,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,85.549,85.5,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,critpt,CritPt,science,f107668c2e5936aa4eb6cdf132d83be520796eeb48c9f62987c9863988ecf4ea,156155c996486e4d40d9c4567ad459bdc7df766c03f9cff33cdca8cd36a4f1ae,ebdb1826134f5d76dcddd272e4fbf9800158d6132a12d69d58d186d1f72ec275,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,12.0,12.0,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,simplebench,SimpleBench,reasoning,a9c651756c08ba8330b8477875e5b834236d116b2df877de6ccaa7f9bb35bc8f,5f784cb9925a911e06cfc09dee368c2a2e7272a9982176217966a51fd3b3abb4,6b0b62de4c90e22ffbf27aa09e82ef448940e448951663f30e2ebd7aa76798b7,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,61.7,61.7,exact,canonical-source,profile_15740d2b88537909e84c3dc7,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,chess-puzzles,Chess Puzzles,games,ba091838f1d5de2f9843b21057ab3d30b159a5ca51db397d9c83aadea2c25d48,f6cc02c57980fdb88867f6baf75a5d5f5049486b3d8889049c9be98df11debf6,1a210fd6ee7b22d9f39554a073841e2ed8d6a6119a7263dadb15eebc9ae20663,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,41.0,41.0,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-10-13T13:36:23.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,arc-agi-1,ARC-AGI-1,abstraction,b766a2a89245241c19f177d46308eef49aaaf481c4f757c4316ed2f540c3d52f,11e3c70819fe67f6ec0f1ccc8ca54e1656c2e96cb19e74dd28fad40b7261020d,9e60576dccc5cb95fc28b9550b2148d78cc1e9210cff2b06fa0049c9379b7c89,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,92.0,92.0,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,swe-bench-verified,SWE-bench Verified,coding-agent,13dd7abcc09c55eaaf075342e4d5ae44e8a312ca707122e997805c0b5ce948f1,3d61a014a031a4aabf68d4fe68fa1b0646765f8148e8554c2c31bbf93ff69451,202f22711959c37116e4d787c4aefd4536edda75ee58e33636241b10b9a6921e,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,82.0,82.0,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-08-28T01:30:56Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,swe-rebench-v2,SWE-rebench v2,coding-agent,4df263ff2ccd91dcde8f47b1aef9c32231f8c94af170d6d99bbc9176f3b84c80,849ff548dfdb194ba82e9190319713ff80cf1c4cb070192eeb282eb8790e95eb,56db8c9f4e6d312b6b00ad167c406430100335479f46b025cf03f8521d17936d,release_332cbf0e979daa9e173de9f4,rolling-v2,2026-05-01/2026-05-15,resolved_rate,percent,true,42.85714285714285,42.86,other-supplemental,supplemental-post-release-window,profile_1c905f9be1ce4940b60cc62b,high,high,true,fixed_effort,,true,2026-05-15T00:00:00Z,fresh,170,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,gso,GSO,coding-agent,f37195ed56d53df0e057c7fb0dd8d97066a9cacfff79338c22f65f13f0594189,afd102f0a9a2b16aaca593dd510bc8554692d06bb419f5a99ed5c088c89eac27,90fc44bc011dd559e0a60b3e81a8069f34dec6660bf522853493ffc7bb164ea0,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,44.12,44.1,exact,canonical-source,profile_1c905f9be1ce4940b60cc62b,high,high,true,fixed_effort,,true,2026-04-27,stale,188,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,frontiercode,FrontierCode 1.1 · Main,coding-agent,5b2286ff66f868bceb66356e28dc3e0b0280a3a13db6a06e4a4868b032f2c19c,ef5b62ad3d0a3c66ed21fce0541efc47239ca09adf7e4adbeefa22ed17e0eaba,cfb3195eafc5e45231888af39f986755ca83f74a702dcb5243a8852e0f4f4942,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,38.54,38.5,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,livecodebench,LiveCodeBench,coding,3c89f2a194433ef144d9badd5eebf607f36ef91f0a3435da21279d531a7c0f59,d7ba21cdfdd82ca40b146def834766d6be069feb7177f1546298c99c1b78b803,1ae76f45f3b5c3c61d867c81bebc9e2bb44abe31201237acc81dca0a65339a25,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,85.073,85.1,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,scicode,SciCode,coding,7c485148594a52cdda3cfcdbba12db3664ee5ed3529722fb18c157ffc21dd49e,80e6acb48785356be6402fd15d442999772dba5848d52741a0adf8635c1b378f,5d8db5a2977ad2d5ef6e4b5cca401b12041207b8e5311c696ee76bdd3ca238d7,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,54.51388888888889,54.5,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,ale-bench,ALE-Bench,coding,3008c748632e2d64379da86146689b5c41fdd6fb6849502791e20fecf75bf8a6,731fc1abd9f3d76e6178794d821fea8ee361ebf72e6213205e103c2c5b4a646d,139cf0ed306fb30eb9c7474fb6fdbcba0ad7b6b87896baf0fe2b823b0e87b094,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1323.05,1323.05,exact,canonical-source,profile_a98b3983e054c2e804d5eadd,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,vibe-code-bench,Vibe Code Bench,coding-agent,13664c031a0dc2a1a7d56e091325dabc277ed2aa9859ccb646a543cff4722d6a,ccbe797fcc5a4830fff69681e64aed555d5b47a08f79596fd175b819adc37b30,88d0f1020e7a012eedb4911b4551cd87edf118d18a31665492e1aeb5530e3926,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,71.003,71.0,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,webdev-arena,Text Arena · Coding,coding,27aa3525de1e923c87f9f37fef84fa79c43e88efcf5b840a4e887a1a029bd01e,799dcbb9fd7def8d7a5d79e4c28716dc073567799ec11261dcec6160a4fca7d1,edef1e738db058717c7980cd576a620849ac7491b3457fcb4d990c8680a7947d,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1557.537773306169,1558,exact,canonical-source,profile_1c905f9be1ce4940b60cc62b,high,high,true,fixed_effort,,true,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,mirrorcode,MirrorCode,coding-agent,91f410454a54bde409181114ff1ee0e5e5af393070d6631d8d40ac4405ef1df4,95cc7aee48bb869eced7f381dfb84a3ceef55d05b3fe1e80a40776302e5b3cc5,10132dca49c37c2b0b2fa9a13ab931265d36626de6efa155ea0e727836011035,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,31.11111111111111,31.1,exact,canonical-source,profile_1c905f9be1ce4940b60cc62b,high,high,true,fixed_effort,,true,2026-08-12T11:43:31.000Z,fresh,80,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,programbench,ProgramBench,coding-agent,d4895fab8f8dcd33a9ae263de9be9d8a4233dff4e7032675993bec61dd52f85b,a8d3f6f1c9c73b2a26abc0db61b33bfa69d2ad7efeb1f4ab7fadcf10d7a1a84e,c0465d9da349e43c8a10e742788027dd69b0140de1f41c46b294b14f03d78e97,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,f5a73224e0a5689f745ff9d536515e9f79962276f7fb29d9ac71a00d99345af6,ffe0947b2353363c651c656f3794605beb9ccc89db067188bc09ed3ec1c8a0ef,b4457f1bd024f1e1eb514ac715e52c72b74052b9ad9095ea552d5f5ce51a718a,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,68.539,68.5,exact,canonical-source,profile_1c905f9be1ce4940b60cc62b,high,high,true,fixed_effort,,true,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,mcp-atlas,MCP-Atlas · Scale,tool-use,0e9cbe3e77fc3ebdd521439bf03f9be7bb98e237f43622ce6ad919b7bb1ac7c0,a2f1b5815887748ab14d7fb69838d7ce4c5fc2519de7594075efb536e6fa7d83,1151055857e22332fe7a552c1c18c8d1c26954d7d8824074dd5d0a72e14745bb,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,79.1,79.1,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-04-08T17:02:16Z,stale,206,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,skillsbench,SkillsBench v1.1,tool-use,ae76e7c3d72abd479e34a03252e1ace9637f404fd18702c3e774542b855620fb,a637cb26046e8e5896dc3aad91f0828a337cc87c9525e2d17db01bd374f5b271,3b4651a644f962cc455e244c75c19a44ac9a846ec9096a32e403b17de579282a,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,61.2,61.2,exact,canonical-source,profile_7135443fd9ef4d3d33bd3e91,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,apex-agents,APEX-Agents,tool-use,cadcd474d61558519bee7bace300c8d4f4e17b2b96fe4b4fee38dd9b077c04c5,1fa31c554139b9d7ca6bc5b0c5510134f700bf2c0071861c77a39795a9f46b42,94ea485b152fe88074253db3ec0e335d447550a40ee7b3563c20392013ba4ada,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,49.2,49.2,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,appworld,AppWorld,tool-use,cf02cbfb2563fa4f2bbca89b8d80a86dd0c02305f14d98aaad6b013623d36369,efd7c4125bb761854b420db11628d2ffb03e05c078f3c082ff47ad18d86373be,a224489dfbe69501c48902cbd5c7303930991bca8273c1a7269f09203a141099,release_3fa145932216c0afcdd9dae6,official-c1f56015cf7c,test-challenge-all,task_goal_completion,percent,true,85.6,85.6,exact,source-specific,profile_82a761bb656269ff1042ddf1,reported,unknown,,unknown,,false,2026-09-02,fresh,60,0,A+,0,A+,0,A+,official-leaderboard,0,official-leaderboard,appworld,AppWorld official leaderboard,https://raw.githubusercontent.com/StonyBrookNLP/appworld-leaderboard/c1f56015cf7c3441ff1933f5bfbec879798d7bbe/experiments/outputs/_leaderboard.json
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,vending-bench-2,Vending-Bench 2,multi-turn-agent,8e829b7294e5d366c01ff2cf79d31a8a7baca058dbdb3a57c5d137d4b6838809,2d10ea0959fb574cc661f181f9c7faa1f4d5b56981d8d22ab91172a85025041c,7e827b07d7108076c5aaa4751be97125bb1c3766c6300a4d9650a45587bcded4,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,10936.763333333334,10936.76,exact,canonical-source,profile_5560a7522edc54c726d81374,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,323743c629845edbec7a9035381315e60b9aa5ff4dbd5dcdff9f91da5d28b00f,1c7a6787ddfe92847cfa7cbdbbc115667eb97870bbe6aca9d23692e9a88b8e92,1e9829b67a89ac1a0e261fbe8c9449e686e2eb6f22d8dd00021d057f1aeaef57,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1361.69,1362,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,osworld-2,OSWorld 2.0,computer-use-agent,cb71ec6d34381acf2255f48bc9ff2340d1b7a25099a10e700e34e37e77495544,9659a22d5f4c7011e9734d8992c0e3ffe5f4360a7bd307dfff8ed8139cbdb390,8324b7d07a58972148ce8c7a94654e6da9a57c9a8f154331c3e283d10e6c2175,release_15c5525fe39f33bc181dfb69,OSWorld 2.0 current leaderboard,binary-accuracy-500-step-budget,binary_accuracy,percent,true,18.2,18.2,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-09-02T09:27:06Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-osworld-2,Epoch AI mirror · OSWorld 2.0,https://epoch.ai/benchmarks/osworld-2
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,014ef7c94f04f917320174ad17ff9530f50f0f4c3295ab45e37593860e5e3c06,a6d4426ce88163ba9c8d81b87dfadef2578753ad280d745b0749ac7578b97155,ce4e31865bc9890925719bee70ecfc72a33c3c49e73aeb5cd7164f1ed47c7997,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,40.20618556701031,40.2,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-07-23,fresh,101,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
anthropic/claude-opus-4-7,Claude Opus 4.7,Anthropic,closed,29,weirdml,WeirdML,coding,ecba4cbfedfc3c475f0041acf4221feb096ddfe4b18836c373d8498702576d0f,5226176b3ccf6c07be121bf86925074cc36781db027422fefd7de120e4728975,b8796d5baa9303df1dafed6e88dba11f1571c9db52d952b1f14c7739aa7ee416,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,75.45,75.5,exact,canonical-source,profile_d21e5b67f44d1fce22bb9f73,max,max,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,epoch-eci,Epoch Capabilities Index,capability-index,4b14741a470588a5a7ca383a1fcc4c2cbeb600523da3e3b65a78c8ca83eceb6a,cdb2641d19d57b7a2cbabc65a55efd084398dca9c25836ee292c818dee2546cc,7b6e5dc0282d24e90e1b1c66c7da5d0e6ed42578e3cff7978a29c559dc387272,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,156.67,156.67,exact,canonical-source,profile_a3d98e32adc7050a02790a04,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,arena-text,Arena Text,preference,f1db6d1043e37def76072cbe89975f2a3e0566407d02a76997a65c36d96f6ad8,95f9431095803d39b2c4e322fe023feabd48ba7d47b2f85c373ea031e72ed72d,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1453.0205808990422,1453,exact,canonical-source,profile_e57bf627bda67d6b38ba2302,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,livebench,LiveBench,general,e4c53da6fb89ac015d075b66560c4e74e8fb74383205831452aefa5e7aaf598d,0ff7e56bfc94de147e78de4016e2ab5b008ba000061b25699df458831e1135dd,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,73.55938095238095,73.56,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,76b498fa8fb71a11a4cbb4151cc43d63e2e6654de77c8209bf34f8232228c9cc,ad93cc86a26aee9067176fb1148207bff02db6b4d13617d096c7527bbd8b2a90,2f6f8d9ce3b94eeeaea3d2b381f04047a9cd2860aa84db51286b1dcd0834e383,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,39.481001,39.5,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,simpleqa-verified,SimpleQA Verified,factuality,7c6c48b2b0a4f32db4a1b85888f68a6594aa30419d93167ee4ae89c843656167,b3f8bed4ee86a272d8d069989d198903f83596e34ce72814a4e210b6074b6ec8,d410be56e6859f4e7081a0cc8ca85150541da72b651a27b3d98ed6bd89c2c16a,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,41.0,41.0,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-08-10T23:44:20.000Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,gpqa-diamond,GPQA Diamond,science,9c9ea0eb34bbad0c29c6a30af3a804dc4d8c48ae2ec534dc3c41081c3d11c524,d60f56ee494004f974ecf74d555b1c2fe2d65f7a7e64ce17b94268b3d924e8fa,555753c3f1f3f2822c8ed4e22fb234fb2adaa2cff3e3ddfd7906dae4e00a9c16,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,91.666,91.67,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,mmlu-pro,MMLU-Pro,knowledge,844bf8007ba19e3d3c767f201ad56d40df24cdeb7e44bf0073106d0c38526fe9,d45ac48176245c45cb0eafe5ecac6f95b9b9b79f970712d765d873f4b02732d6,da56b1fdaa09eec0b14f50dbf76c5fade490cfa3db7dd410766d09eab63928a9,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.036,86.0,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,58b3c5c724390b300f1661758308331e83da4844b871a4058d316c8110140038,210e8309abd8f562a3f5772f7e33d8192f8861086126618bb1760cebdca056a5,0308d85f90369d2eee9ad2f5f23e877c1b2ad27b956eee9188603f10586fd690,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,98.33333333333333,98.3,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-07-09T13:36:49.000Z,fresh,114,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,frontiermath,FrontierMath T1–3,mathematics,d1f5f2001b6880dac8a273b67dabe365b96e3af2bef509a626b26378fe50acde,3c7e725ea4ee8cf88a942f1305d3d52be3f05835498432f69529b3f59d905bc4,4b8b8ad290dc029c8d200cb56c5ff20f2f2dad01315a27e7ec1afd440109a271,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,82.10526315789474,82.1,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-07-09T14:18:35.000Z,fresh,114,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,arc-agi-2,ARC-AGI-2,abstraction,efb2ce103eb9f40c3a280c79b54edcb51c99668678df0ff56b89abdfdab596d4,a53a3ab522f1aa955abcf6ab72e0599cf9660b953160f129dff83f867f2f976f,f1bd525dceb8715c61fcc62b3b1067943be8d659f1b593e5d8d640e1fb897110,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,59.583333333333336,60,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,frontiermath-tier-4,FrontierMath T4,mathematics,9e752b048490fefe9a253eb79be6f6aff6b44fd18656120d3da1821e1c670ff5,1ea1dd3b0da02a5865046bd630afc7e699f94daf966989c5395cc7266d8b9ec2,940a93eb7d331a9efe6e8ebe8fca4d3617f421d867cc2d2dab2b94a82cf24e2f,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,60.97560975609756,61.0,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-07-09T14:18:35.000Z,fresh,114,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,mmmu-pro,MMMU-Pro,multimodal,244ffcd1ccca476218c933ea49b6e2653e1022e5bb67c3bc2ed4fe3b9f0a5497,ac3e2c671fef94e523b036154fbc8f891bcbdcb0090236f9c9bcb81c0fd847f1,a741782697281ded2df15e8b50e7846d3426a3b9f6525d9429dedffad330c9ef,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,85.029,85.0,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,critpt,CritPt,science,2b6c30aab4c3e0e8310013bc6d8d32f05e8b30a205906319eff4a11e0d003528,42f806c6784f5c54e532c66303eff8487fdcfd310ef0fd6bfe66915fa8dd8a00,662f76f2ffc17c94760e16b2d64e236163fc20f32c09920d6a6beb7bbf320099,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,20.571428571428598,20.6,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,simplebench,SimpleBench,reasoning,f93627841d21fd1aceb3b814859107a685b4aad286cda30dde52b157a49b1c04,7e8d6432b16f5a6d5a9c3f8475a5a5023ce0d84d18a96a86ca9894ffbac3f717,1648b670dd9e109a8ca13ae1a8d575f070c45134efc8688a0b7f8d6e92abd9cf,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,46.8,46.8,exact,canonical-source,profile_e57bf627bda67d6b38ba2302,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,chess-puzzles,Chess Puzzles,games,fdd460cc0589840dd956283272d34f122fb769c11722b3575fe5dd4671fd1b97,2023ad3153a714ec07d521a4692c4b6ba28d278ffb0fd6f83cb1b28f040ca048,0228e713ccc4c27cdcbb407447229c9c7e60289d0cec66b446152ea13c9446ab,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,43.0,43.0,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-10-13T13:53:45.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,arc-agi-1,ARC-AGI-1,abstraction,b8fff8a7304ff8afa12c298a1209f886a2974b684c0a5175346541d5fc95a843,939508c80b4ef2cd545843da674444e2b6df7ab076714189b206c3e3cbee2ba3,833b5eb9853713ace12830a8bb02a8f907235c2ed7576989e479d4f57ffbb594,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,88.0,88.0,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,swe-bench-verified,SWE-bench Verified,coding-agent,94c1ee6bcdeef6d06594447ac4c5e062f2af9583fbd865cdf6e8ee4158e67112,0c1c9d99bc233dba16e3d66b9f86ce29d6d72cf491e8b39340d47a5467533a16,edf0910f1683a1c6b396f030d187ef7ebab83f72bef7c33ac91d82991e8d01e9,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,93.0,93.0,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,swe-rebench-v2,SWE-rebench v2,coding-agent,5f3ea02a4a93099996523bccdf15507986235643d6f8618a5cf22be9d7d9e07e,d6102f84c0897c12091cbef601a473d78c6f0d573d85b911dce232aea03f70d6,4bbed7a27d62a359466c297ea94c4763d7e66439d93473a4ab91adb70a3a02d3,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,43.6036036036036,43.60,contamination-warning,contamination-risk,profile_d00d7de512e3af291d4a5480,medium,medium,true,fixed_effort,,false,2026-07-01T00:00:00Z,fresh,123,0,A+,0,A+,7,C,potential-contamination,2,potential-contamination,swe-rebench,SWE-rebench,https://swe-rebench.com/
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,deepswe,DeepSWE,coding-agent,c2c4746f44933525a55f0f6be757ec250707ca5f933a1407f614ae238ed6ff99,b05c3659f52b5854b8ae228d95f895551d8cb65c05c044570ea2f16fd45302fc,b4b94369640b3c39e71eabb0fdb8f097be0b63c1b2cab8ccc91fc7e330628e9d,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,67.1875,67.2,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,cursorbench,CursorBench,coding-agent,76d31ab04704bc1c5b549aa5cbd945397c5f7111badbe801b9f1bdf86de95809,d6468189aae46b5373e9cbbfb7d04f726fff4b39a42cf0d7746582da3b6cf0cc,620792494e9d2469a68dd2138d94612d534805260e1d9384cb95369af7f5ff37,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,35.9,35.9,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,frontiercode,FrontierCode 1.1 · Main,coding-agent,1ee67dec805121ebb889a512a9595ff5408439571b6d8f38089c6cb17bad497a,02a4abb16f933efdc6499c28c3a1cccbd4040332848e285b68f2d577bdc5e34e,97f8a238f9c83cde2459efea5678c908db5f00a8835b0048d51e297bd9df0c56,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,39.81,39.8,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,ioi,IOI,coding,873165a97be933f9498b90a17a72d57f0380ccb1ad30b10f133c74871618bea8,1fa719d755dba2df7ed9a72c558fb0fba28f1ff845c0b6a3cd378cb8d3b744aa,a868d2da333e27cd8216cfea65643a42b41e5fc16a7076219dfdef899cacf073,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,61.778,61.8,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-09-14T23:34:42Z,fresh,47,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,scicode,SciCode,coding,2b9b6814f4a40e73ae944590f3aacfeaa3f6fa6ab940c4646aaac8dac1ca6588,075482b31daf64c2fce092cc98bd700158c2fcdee64f3e5a4006cb06843c3f1b,5ca8a356da506d1968447df78690233c30bcb75f017227ca34ddc2a4ae210539,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,53.587962962963,53.6,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,ale-bench,ALE-Bench,coding,bbd6f496c02d0d5f8c1d031ad531350af0bf1d28228c9cca01108bd489e4efb5,8dd1fdb4e547098fe8445c99c588b8d540ebac746e198a37941b63283c5b5359,8970e59b1071083f0b89273e65b29adb3f889938ffb6a4156d776dbbfcfbbafd,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1667.4,1667.40,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,vibe-code-bench,Vibe Code Bench,coding-agent,df872bb9130fe83a6194d3ef3c34a39cc7fef31ae923baa171391f3a4fd4790b,0b743b31febea04ab2097da57e1be0564962eafacaedff20957cc589ea4d9511,e448637b07ebd58d1217a41467338dd4a7fff6f7b76f4b2169262ecaa088471d,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,77.055,77.1,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,webdev-arena,Text Arena · Coding,coding,5749f8199dd24eec699c6b80841556f9dc7db51a2934953e41b9a71db753d7f7,3b2645cc254f993053bf45d2aff01248963c2cdb9e30f8a048157f6f9a5f5a47,8d5a80386a650bc2342c19dfcc983bf76f89db6c49e9bbfe3c5e6c497e75057d,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1518.9116265563573,1519,exact,canonical-source,profile_e57bf627bda67d6b38ba2302,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,programbench,ProgramBench,coding-agent,907667e1b8148fe8542b4d28cb51f948e9f16203391917a1f24f7b498983fc3a,6180c2b6dc7792303b9dfbf6d41fa31d30db8bab3c313ae48ee5b04261b8d8a8,ea2449a6fbe41b0fea8670f14e3e8982ba381fb1c692faeb9f9418d580610616,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,weirdml-v3,WeirdML v3,coding-agent,0df931576d178f7ce64ee627447b54c544b2d06922d0a86466624e5dff9ca947,d6062913e3fa4adf1a7e4b9bd59d1b1196ee9bb7e35cd42d396b88ef855a0d78,ef82937aadfe5361b9faaf7035d33a37fe12630e408a5bed2de566f54bc60bcd,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,5.1579522727272735,5.2,exact,canonical-source,profile_e57bf627bda67d6b38ba2302,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,005700f5a095c8e37e6a15c461aab437477957641bb2c200d9c41e9e14f7fc13,4239a196534b1d21b563bd83ba515761c32b865839cd3c3f86f3f591509362af,ee018f9481251503dbfb1f87ad00f3d1122406f5386764952c75e2f6856d2a21,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,79.026,79.0,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,7aa49fbb948e5a56b146ac610392005b325a576228924e6c4ed90d883dbe3f11,78bc911f8c6775ffd2e5641579e71417f0a5bf41e474a2d0dd4d11d1ea99adac,e8f29ec103b9c670b74c48456c127f4e5ef4c04756b3ccc4587ebb4a4128b22c,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,11.616,11.6,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,sre-bench,SRE Bench,terminal-agent,b63d2a3461f7a0f97bb28fcadeb3531b6277bacfb96be310fe86b26be556eb1d,6c5835b29d26372fdedfd5a64db440ffc467822974d1deac7a9e8d6f54e25c16,731dc3293a276ff4a4e4f007852bef6fe69531edbcfe0b2af8192e3c94683e87,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,1.527,1.5,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,b4ac1aa4f5b02310a9a310e0e58c4864c18f3f4d7f02142bfd10e21b04fab2c3,c35590786701e97bb0997ee8421949356d8e9adf38e44d4f5d0d7e30f6758cad,2f0500d763c3ca201f3996008fa7227c097c74f25144824b446e5861f5e4bba1,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,31.134021,31.1,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,apex-agents,APEX-Agents,tool-use,5ee9253ff5f3fd0d850634faade9036fcb5107ba80e8b4295f0ca835c7402f05,05c84099bd6aa60c3681b5bfc4ab8970335e649e8cd95156ffc85bdec16c107f,9f2153b6c7e2764fa0413e9d6b9456680fd759f2d7c9f834cad0639ef8eee045,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,43.0,43.0,exact,canonical-source,profile_9179a50784f1edd95e286993,reported,unknown,,unknown,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,appworld,AppWorld,tool-use,cb4146b5c08b1dccee907b1c3991f2eaf2dd48142fc33471381bc2f0a19717a4,45c45e459ccf4e11c60e8a2f4bdae7f5c11801921418694f5cfa13b1f3d5a78a,ef1dc06cb4f5d702032de3e7246e740e9fdbc27cb185f697fca3cb42f4d74a81,release_3fa145932216c0afcdd9dae6,official-c1f56015cf7c,test-challenge-all,task_goal_completion,percent,true,73.4,73.4,exact,source-specific,profile_ae9c2d1f81bc6fd257f16a98,reported,unknown,,unknown,,false,2026-08-08,fresh,85,0,A+,0,A+,0,A+,official-leaderboard,0,official-leaderboard,appworld,AppWorld official leaderboard,https://raw.githubusercontent.com/StonyBrookNLP/appworld-leaderboard/c1f56015cf7c3441ff1933f5bfbec879798d7bbe/experiments/outputs/_leaderboard.json
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,vending-bench-2,Vending-Bench 2,multi-turn-agent,34b0d9de58384d09f47f0c25b7d005c926c65952bb4818bb08950f7ba561c4ba,4fb9eff4daf42ceef2c9b3b08a07bef60a6f8f9dca0fc8a730463906d93c0678,93ba9825f0574e5d080f2afb2c579a469251701b6b0e8ec17f61ee00ebb8a257,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,4094.7120000000004,4094.71,exact,canonical-source,profile_e08847800071627e0e69db55,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,728674e089a5ff91ea75ddf5f649a6f656dd00a915d6437e9176cf15d0a72158,e0c5c50a92da7dab8943cac05900188d3812f6df566016e58774e1cbb88d1224,39898c15322729386fcb38382ced24f5855b59ce9f0c8805f4818bf0c69be5d5,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1468.24,1468,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,terminal-bench-3,Terminal-Bench 3.0,terminal-agent,81c952549a478f4ad4ffe2a7859898dac4e507f5f3357e10cce51cb3e3ed5ae2,dd0c64f78a79ef68fd9d3c1b6e28961c2be02ec9296d9f2456ed0633100e5b63,7de4f97a12e9399bd298fc0584c86bda697ffff8673185d4fc32c4ed36b239d2,release_ceb36f25882d8cea2c9d084a,3.0 · b936386b-4afd-4ad7-aa7f-6b52bcab9853,official-74-task-leaderboard,accuracy,percent,true,14.32,14.32,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-08-12T21:39:34.661007+00:00,fresh,80,0,A+,0,A+,0,A+,official-leaderboard,4,official-leaderboard,terminal-bench-3,Terminal-Bench 3.0,https://ofhuhcpkvzjlejydnvyd.supabase.co/functions/v1/leaderboard-read
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,5a59647f4eaaab3ef9603702542aa95a9544f1f0275d7736c6321dd19a2dd8e8,e588736461705de9daad4c8479d0c9d14f376246691e72844af054b2ae0408a2,33bb3ae1c79060a290c4873469d7686da457b835b2ec1c52384e9e99af1b6c7d,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,0.0,0.0,exact,canonical-source,profile_1b694cc32a697b0749e4e7f5,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
openai/gpt-5-6-luna,GPT 5.6 Luna,OpenAI,closed,30,weirdml,WeirdML,coding,52698e31d797c183a2dc7234c3ea1aa85b178bc0cfca676a4edc509c86574b73,53ce27ba152883d7fdf314f0b1e8274b4cf1d38bef604798157102b81eedddfe,4c42f574a3937ca56a96171015b4b7db3075113a478e4f98ae2cbe31f2e26dbf,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,60.86,60.9,exact,canonical-source,profile_0bc8d2a65f5e960dc81d9df3,high,high,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
openai/gpt-5-2-pro,GPT 5.2 Pro,OpenAI,closed,31,epoch-eci,Epoch Capabilities Index,capability-index,ab3d409e91cbe15f9f354c0b1fe267d37e030d064940634be3a1f242861e4227,474ebbe8d646d2f23f2254b4871abb28bcc2da0907b31a051aa73d20bc266b04,10c454214b47e4038d955a2827067dedc2b2e6acf443358d9b2527f62ccaeecd,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,156.54,156.54,exact,canonical-source,profile_fd182e68dd93da34ca6196d3,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-2-pro,GPT 5.2 Pro,OpenAI,closed,31,frontiermath,FrontierMath T1–3,mathematics,7e8bcd2e04110e2bd9d0ed791519476e181730d2c68e9f8b0adbf09566a7c914,975675bf9249a52966cb429e8bb9448132001054c1c7ebc86abc723f2ef54994,8c74b08a97c0383941cf14c84ca5818f0c07fa819359022155073dcd71799c6c,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,74.0,74.0,exact,canonical-source,profile_95ba440b37e9b02e11c009b3,xhigh,xhigh,true,fixed_effort,,false,2026-06-13T02:37:55.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-5-2-pro,GPT 5.2 Pro,OpenAI,closed,31,arc-agi-2,ARC-AGI-2,abstraction,42c03e91a948256e3da8779b8d3357d104ba43a9e1e8fe47fd846c36f3468546,913d8e0f7f9bb89c0304dda4ee6c8d412e82bd6f62f36aecc12fc1562f32cdcf,2f85e349795acf0d68e5ae22a7718ee829611fcb56986f06c43659bda088f931,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,54.16,54,exact,canonical-source,profile_2fa48c668af270933840111e,high,high,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-5-2-pro,GPT 5.2 Pro,OpenAI,closed,31,frontiermath-tier-4,FrontierMath T4,mathematics,6b59027c5c5c8e0ce637cf8479fa03f9a96fbc090383a8826ebdfb7b5368f5e8,a37a83985293568bd92847df77570526b689a68fcfc72742c147aa03edf7937c,97480455bd56f5d849116086cfed4deff8497a6a66c499469ade8dfd31e3d42a,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,46.0,46.0,exact,canonical-source,profile_95ba440b37e9b02e11c009b3,xhigh,xhigh,true,fixed_effort,,false,2026-06-13T02:37:55.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-5-2-pro,GPT 5.2 Pro,OpenAI,closed,31,simplebench,SimpleBench,reasoning,a9533bb9329263775a7ef5d5a8fa24722872aa189a4d18a92c4eb58f795c7283,ceb1fd7f3a64074c3d4f7d31bacbc8cfacfb8a41c5fdf4cfcf02dea6a65a6bcc,641e8efe24c5d7264e335e4f1b1372035aaed643f9a70ef1a1b65a599e3f5883,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,57.4,57.4,exact,canonical-source,profile_f84193b5a06372776da088aa,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
openai/gpt-5-2-pro,GPT 5.2 Pro,OpenAI,closed,31,chess-puzzles,Chess Puzzles,games,8aa97771b96d041a81a42bc87778ffe706e841379d2e5d33450a0cfe15804641,cbf43f1784c146f7b01250d6340fcfc5bd880891fb11333288f8a64f17bbf535,71a59035921fddc10f2ca3cbb3a31faec8e62b04c8211cdc0f33be58d88ef8b7,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,60.0,60.0,exact,canonical-source,profile_95ba440b37e9b02e11c009b3,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T09:56:34.000Z,fresh,17,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-5-2-pro,GPT 5.2 Pro,OpenAI,closed,31,arc-agi-1,ARC-AGI-1,abstraction,b10788223355f6a259a99b751e8f7db5cf703f36f3b713da73588e4901b7dd5e,6cfaf4e7321da28395fdb28059d079239c207fc1e33328148f7ca9a61e653b7e,254d25a57b3cc806c85d610086ba76a6ff094450b15629846f2053fd880cae07,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,90.5,90.5,exact,canonical-source,profile_95ba440b37e9b02e11c009b3,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,epoch-eci,Epoch Capabilities Index,capability-index,b99dc7faa271bd21a71b97976c4c83e67044b78e2121664fe6265cfea6bdcb41,709ae3ab7206139a98e5ca2ce0365c40895d70b66a4c841e0e2f87a12a1ff818,ac68c2a4f63454d624b58a86d09e59ffb5e262ee2a590b2d1f2a57974509741e,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,156.24,156.24,exact,canonical-source,profile_164cbc78e507602d5ae2f029,xhigh,xhigh,true,fixed_effort,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,225a9d70f62f0df59bfcf42bfb25c1a8a1c01d0c214391d46d11f253d3ef6031,f0b30eb46b8c66fca5e07d007475ac7e96d245d4af9f511c681e615ad6798a27,6f28b4845c47581aee73763fa418d85c5501ed8805b5335b093e3250ed66ff9f,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,33.554563,33.6,exact,canonical-source,profile_fb817aa80071f23551b9a00e,reported,unknown,,unknown,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,e565a4f8ff08338be3bfad603f79baa2c42549e8b2a5b1fe0da7288449e6afe7,0e5e73134c779a80096c16d8581eadc76f25ca474ab4c4056a26285105ed68a2,668a953aff1136f3db12265cd018cc039a645e3612b87ed0bee2099776aca051,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,43.095459,43.1,exact,canonical-source,profile_fb817aa80071f23551b9a00e,reported,unknown,,unknown,,false,2026-09-17T14:33:50Z,fresh,44,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,simpleqa-verified,SimpleQA Verified,factuality,8716b12b0710a7dc2e737d9142dbf3de6efafd64879184123405f483599b913e,920c44e510ba43543ab94756a5a1f079cacd010b48cf179d1b913eca8e63f375,7cad23dc5e0345efa51a17c61f2b76c4524b68b9bf3f9a669774a1e64a094e38,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,47.3,47.3,exact,canonical-source,profile_164cbc78e507602d5ae2f029,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T18:04:58.000Z,fresh,59,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,gpqa-diamond,GPQA Diamond,science,914fde793f0d3c6e125a6a3fd4edb44534f7d8a8ea06b63f1f34bb748c2687ac,9c713899881c03f9cbe097a0aa8f3386c8dfab8a4f5c4f48b4597bde519f463d,bb6eddc715728ff54e0ab07a9085d24798298d07221648218b40613d1993e431,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,92.828283,92.8,other-supplemental,supplemental-source,profile_fb817aa80071f23551b9a00e,reported,unknown,,unknown,,false,2026-09-17T14:33:50Z,fresh,44,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,1367c717928a0a5693d4ba2784e4a4fe895117713568e67d5f24e7eab251bd4a,602f2e8909bb0aa4757a6fda32495dfee6fb73eb5509d436b08873a4f545f2cd,540abe7eb1643f4b49284343fed302469c7b111c03a249b3fb97f0619b2a8dc1,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,100.0,100.0,exact,canonical-source,profile_164cbc78e507602d5ae2f029,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T18:04:58.000Z,fresh,59,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,frontiermath,FrontierMath T1–3,mathematics,0c3fdc568eb4eef520105eaa577882e21c26b7056dbd9167e61fe0939b729d7c,4cd3cf47b84dfeeff0e6f4623b7a7f791081583e25cb7c4a247ba44387085757,15eb55a29c8b986596dfb10383777044168822a0790707a85eea498f76ca2d71,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,65.6140350877193,65.6,exact,canonical-source,profile_164cbc78e507602d5ae2f029,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T18:04:58.000Z,fresh,59,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,frontiermath-tier-4,FrontierMath T4,mathematics,9fdd9a529c79db11e34fbf24126a8ba8b8032e05846620ed5794991ba804aa3c,af3ae58a58e2a08c0cc18fd617b3295f68192f73b153245002d402cc0f8f016e,3929c6c9d56ad502dc10a72e084c4ff9e06c385d940b2b5bfe9df5f695a338e0,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,34.146341463414636,34.1,exact,canonical-source,profile_164cbc78e507602d5ae2f029,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T18:04:58.000Z,fresh,59,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,mmmu-pro,MMMU-Pro,multimodal,890e4aea4fcc48a7396a21a9e416781054ae6df25628f6ddbd8f554d783a793c,8f4bc72538def87515495b8a905a428c460dd26a705e023bedb92970e18f0fe5,23b3f8c4e2a454fadaa2d7f0dcc88335ddb5da53419649f9230e9ad5dcfbc92f,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,82.774566,82.8,other-supplemental,supplemental-source,profile_fb817aa80071f23551b9a00e,reported,unknown,,unknown,,false,2026-09-17T14:33:50Z,fresh,44,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,critpt,CritPt,science,dcb72eecd8e01e6fd08dacac77b2308911d4e6bd0ddd75adc9a777cfb795a54f,2619688410dbd05b702bac36abfc595c3479060fbf5bd3287c4effb79a174282,a0ebb77a59f59ccaa16ca5ae119f6bd278c6d21b5a4d079aef7548607a6fa80f,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,17.714285714285698,17.7,exact,canonical-source,profile_f1de9baf1c06fe65d3662dc1,max,max,true,fixed_effort,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,chess-puzzles,Chess Puzzles,games,30cfec14fb0bd8b1e05c5fa8501e8b67d98e3d37ed9a55d080ed9e7c7a62e119,8b8c4d6cb97e8f99d82e0f45bcdd147d98a2ae2cb85893938db967fcaecf6590,310d7c1fe372ef72f8287ff102420bf0ac6d43b3c2ea58b8af66fa51b5189ec4,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,37.0,37.0,exact,canonical-source,profile_164cbc78e507602d5ae2f029,xhigh,xhigh,true,fixed_effort,,false,2026-10-13T13:00:11.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,scicode,SciCode,coding,ef77e647191d3cd29a61654e8736cb39fd84d256355ced0b4f50900fd3f0f05c,bbf96a30d0b88fcf2041215890de357504d843e5927c9927374c73667a7d01d7,9d299d21e0ff661f498a2b851f71bc27ba9719c7c26e255b52bfa2384e3ee0cf,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,52.0833333333333,52.1,exact,canonical-source,profile_f1de9baf1c06fe65d3662dc1,max,max,true,fixed_effort,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,webdev-arena,Text Arena · Coding,coding,bfe73ac9e75a256340ff0c8db1b9bff739e5f541e30ad403c4bb167f7877df8a,c81654f696999284ab5382ddd172853b313fab5882cf3551cc0639e6be053fd7,929f644a025e713d598d602ffc1b9d61ee52a5d0c321036bed7caee267ec1c08,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1675.9355584770149,1676,exact,canonical-source,profile_30a83a958902d8055b81e819,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,mirrorcode,MirrorCode,coding-agent,30ce4f06f00bf84e2b2ad39e1076aeae08dd56f53724756637b98d50ff61abe0,bd1044e5688049de45b40ef0746674ce7dda87c19927169c8005704234861216,70d000462fad855cc44a1c07f1350b4a6d456ec0d4b9dda896cea5d6d2c369e5,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,47.77777777777778,47.8,exact,canonical-source,profile_164cbc78e507602d5ae2f029,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T16:48:52.000Z,fresh,10,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,0780d25cbef678a3dcfcc3ecd811be0f7b011ee12ea0b40859f535fdb0b3ac53,85967bf66bb7e64c428302a2886f5244573eeeae20ac13292c26a42ea67f4f5a,8200447c61e9b67007e665430d3d50d523b15381f508bb3f41ec0bad61d2618b,release_c53c830678318dc1763793c8,Terminal-Bench 2.1 (Artificial Analysis),terminus-2-89-tasks,pass_at_1,percent,true,88.764045,88.8,other-supplemental,supplemental-source,profile_fb817aa80071f23551b9a00e,reported,unknown,,unknown,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,461a800c09c7a565aaff205a884b5d0803bb2b21d179bdfca32adeff783b8bf1,5d617ed784f6edd62449b96293014c3102cefda35bb313c1ca6c018b4d6aa1d7,89b9cf67f38aaa76247bbee6f2b9e496dffef731190faaa872210e621ae2e005,release_b7d9ab744ffac0cb856e114d,Terminal-Bench 4.0 (Artificial Analysis),mini-swe-agent-66-tasks-500-steps,pass_at_1,percent,true,38.888889,38.9,other-supplemental,supplemental-source,profile_fb817aa80071f23551b9a00e,reported,unknown,,unknown,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,b871d0749609b783a8e2c0a176b5427dc6c4227118fca864ff54de68db0cbe22,37d9c31a0abc4b782ce9ea8813476d839ac1843679594b37ef242bef4121f4b4,59a0ac53ab5cac2d61eb20995da6e7bb12016bac50f745ad6ed1953dc2f26fde,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,47.835052,47.8,exact,canonical-source,profile_fb817aa80071f23551b9a00e,reported,unknown,,unknown,,false,2026-09-17T14:33:50Z,fresh,44,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-max-0902,Qwen 3.8 Max (0902),Alibaba,unknown,32,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,e8718935978739c9220123453bd7ed5acfb516f93bad23fc56e7dc7623fabdb5,0034436775a9c287cd3d170175ddca51931ecf6324beb0f4342142ad681670cc,d55f72ce119c2dc7e07391184d28c792e9618eb970e0b9d0e23ec89ac8e39ace,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1673.9,1674,exact,canonical-source,profile_fb817aa80071f23551b9a00e,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
zai/glm-5-3,GLM 5.3,Z.ai,open,33,epoch-eci,Epoch Capabilities Index,capability-index,b2b01166e950ca5dfa496fe141fe783fc2615444f4971faa95a41841a5141c73,ea5b24f25b21b8dc7974f7dbe8969a541ca660b26d5c7ef5a238b7cfbc278d80,fec0e71e09e52e528f088c2f0d829a3dd273550d7b9f914f88bdf9e31ec32914,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,156.17,156.17,exact,canonical-source,profile_35a37f6bd374e637bd675128,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
zai/glm-5-3,GLM 5.3,Z.ai,open,33,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,591933a4eb7e109ae720c3cb943cc72fe06c8362c746ad3b83c450cabe8fef3f,2bda056ebecbb69f32258058aad6f74e2754dc1ffec9abea8878765a06052c2c,d7566361e184617ecaa4c04a7643627fd68b08138e6c6ac9ff36b0984d316d5e,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,31.653382,31.7,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
zai/glm-5-3,GLM 5.3,Z.ai,open,33,arena-text,Arena Text,preference,e3a2458690dd12ee9ece72d3ed27bd8fef456b914db26748cf4f5af7ab7dd0c7,e4998e8309524c25c0004c046baa8273cf384d74ed0007f1efabae7c857803e9,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1480.5693804560872,1481,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
zai/glm-5-3,GLM 5.3,Z.ai,open,33,livebench,LiveBench,general,03e2ccdd24aa45e6b66d3d94b14a7af0d3184695de38ed7669c44f1864513f47,124c903d3cc35c73c2c9c497aac27a0d0f90f63f4ce8f7ffb9f6ff7315f4a9fb,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,76.13780952380952,76.14,exact,canonical-source,profile_59194fc28b0dddffb241f8c7,reported,unknown,,unknown,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
zai/glm-5-3,GLM 5.3,Z.ai,open,33,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,e5577495eb7297530e08f01c75b0646c9f274ac11cdb1b98ebba42219c92a481,ba3d4704474cb1459dbd22916e347f992876ab96c9d117dc9bd8e5edb70dd2f8,85de386bb5b9e508e9153a921e304c475ad7ad4f383514d2db7e317cbe2264ed,release_fa9e908ac04b067eb00182d4,HLE official leaderboard,published-leaderboard,accuracy,percent,true,62.5,62.5,other-supplemental,supplemental-source,profile_1e27a40ca87a4086efefafd6,reported,unknown,,unknown,,false,2026-08-29T23:52:43Z,fresh,63,1,A,1,A,5,C,self-reported,6,self-reported,hle,Humanity's Last Exam,https://huggingface.co/api/datasets/cais/hle/leaderboard?limit=200
zai/glm-5-3,GLM 5.3,Z.ai,open,33,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,417e1cc4f5ddcf68d8f57fb975bde04e48448e47f770e3dd5de402326b9df033,b320325eb1c51918cbc5b2c12b899f415e1c5c4dd622cc818235737cd9623a2a,e9754924490d84f96a0dc72824bf9e67b6fe0ab7c8a718f9c17f26bffae726b9,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,42.261353,42.3,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
zai/glm-5-3,GLM 5.3,Z.ai,open,33,simpleqa-verified,SimpleQA Verified,factuality,bad4f506905b0fbe8f4b42ea1d85767021786c7579e953a4ac95cfdc9efd34b9,f2f47fb20f09672a09e70069e52699faf152c4e1ba36f0dca35294c793c1d88f,be976fb14801559fcbcf0a3c3d24dd863d841e7563f53649aa5b2d9b1f5d7f06,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,41.0,41.0,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-08-28T17:00:06.000Z,fresh,64,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
zai/glm-5-3,GLM 5.3,Z.ai,open,33,gpqa-diamond,GPQA Diamond,science,1e7d0c6260914d9bce1545ba9cd7c661b7875a71c920eb8ea574c19ec2c99b67,582853ea98d21d304a03873332326f1e240113ddecf27a9aff8a9159cd3511fa,8b2d4366d6de0d0b8d9a61ae129336054e62c571c9c028b715847b878e9f7f5e,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,88.132,88.13,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
zai/glm-5-3,GLM 5.3,Z.ai,open,33,mmlu-pro,MMLU-Pro,knowledge,b5f93034b53c3d6e6b91f357e22bf612f626fc093ec38c326c3b5063e3afd956,28e687892d7531ed132496181e5f6dca96bb44b750285171979ca6827adf8e5c,508a7aaccefb4d82390e1a87a00f6816f68df82c84c42b41e7220e9a567c2dd7,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.77,86.8,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
zai/glm-5-3,GLM 5.3,Z.ai,open,33,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,3f0122c99d2458a32998dda62917b857df3129338c3d941a86f467b66bf0d0ef,33a8a38255bd04cb6ba24d67afcafec3947be9dfc40bbcf607b7c2a3c9e314ee,4e8cb847c4372d821424c8e84d071e4d9eddfa79832e97f96e4beb381d508ce9,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,91.11111111111111,91.1,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-08-24T19:44:18.000Z,fresh,68,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
zai/glm-5-3,GLM 5.3,Z.ai,open,33,frontiermath,FrontierMath T1–3,mathematics,8087f469cabd2dae072eb7c0f00fd489b88b09c298b98d0cc5db2bf918af047d,7764a9d85197e18373753df8f9b45b4420481162dd78b3dd0c3e093298e698ff,baa8b449e7584a3a04da45802bc5a09fb052cc1712837ea86e91e067754922a0,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,68.77192982456141,68.8,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-08-25T14:20:20.000Z,fresh,67,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
zai/glm-5-3,GLM 5.3,Z.ai,open,33,frontiermath-tier-4,FrontierMath T4,mathematics,efbee2653702d66c1f7d35345c87bd00bee592e8c4440b1f710bace3a181defb,7cd3ae737ed59f10025e74c70dfe3bdc9bc1988e8b254d39a6d13a7c1188ca2b,7bb80b144278a2bde3d74797f51778f83a6e103dec6bf21308a03e696749ae13,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,29.268292682926827,29.3,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-08-25T14:20:20.000Z,fresh,67,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
zai/glm-5-3,GLM 5.3,Z.ai,open,33,critpt,CritPt,science,af994143f85ba361285a16c3a8f5cb6fd6853fbc19b4787647b4590b3aefb373,be50832f36573a7097c03565a6f235dad9a80b274bb99be3877b1e2c4226c56d,398619199077233d6b9028afbc063cd29b8646598ef0c5d52f409695f4492988,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,19.1428571428571,19.1,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
zai/glm-5-3,GLM 5.3,Z.ai,open,33,chess-puzzles,Chess Puzzles,games,6b9daab7931a606a56ba1734ccb97d51b7b207674f3623faa571c2efc28adf9c,810d886a602ceca7b2b8a70d0c260f1a22eecd150fa4c3ae38c6c413f587dc60,4272568acc80114e6a26c9f1f042cc432b86e263370ada55e55018b2cad07dcb,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,37.0,37.0,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-13T21:22:18.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
zai/glm-5-3,GLM 5.3,Z.ai,open,33,swe-bench-verified,SWE-bench Verified,coding-agent,c31f05ad91ca7e97f4da619d2bb1f4b23cd8e75a7b53aa2cc39a9e7a09bd5997,e0f3054018a6db156a80dfae2630b60d7acb28f615cfb57d8c847a39a86e0dce,6af5bb58c3385d3d7246e24eafeda47bb66625ea07efdb3363e7e917270f2130,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,95.4,95.4,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
zai/glm-5-3,GLM 5.3,Z.ai,open,33,deepswe,DeepSWE,coding-agent,1ede9aebd761ccea03f90e5410016eb80c9c6caabda5bbd03d56bfccf473e609,41f75462411b9de0ecb0ba1f07a859047626474a2948fe71336e06b9b58323a7,76a7e4fafabe3bbe3146b6c402e8d8e6542f2b5390eb800eeb16eb81377c59dd,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,68.95787139689578,69.0,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
zai/glm-5-3,GLM 5.3,Z.ai,open,33,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,5e85143dfb7ed9e60dc236cce097894aa8a8b4aabb9b979420b82a373855722b,afb95ecd349bb03943244e4e88ef5cb4ae6e9085c63d7b650d520ae96662bca4,8c7487ef8e1ae1c8c73d12500501e5a80cdcd6d326a9901281b6953b02f8ffb7,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,95.6,95.6,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-07T01:44:58Z,fresh,25,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
zai/glm-5-3,GLM 5.3,Z.ai,open,33,cursorbench,CursorBench,coding-agent,73d5e870964f7ba26f9e76904fb9e6049018d056cf80c341d247b4008b62171a,a8acd540bf973d558e97adc2d8c681fe7d565eed676dad2ae5d6007da4159b73,8eb1cb1268f25dafea962b50eb8093ad40a0bac8d9ca7d996701b49e0e99d0d5,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,42.6,42.6,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-08T22:23:11Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
zai/glm-5-3,GLM 5.3,Z.ai,open,33,frontiercode,FrontierCode 1.1 · Main,coding-agent,e6d3508da7b4b1c8570279590da0dc8fd116a154b1ef14a57688d3b4433d903a,cac1c8f52c2b5ec92ea2cda1953d216a82ab7bd6ca817134217790e7c573cea5,a98c3ab7e44c58a2e935e799cef17cb35518538038c6947533bbe5a02312f726,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,40.14,40.1,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-28T21:25:01Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
zai/glm-5-3,GLM 5.3,Z.ai,open,33,ioi,IOI,coding,949d703253e07d1fe4a978a0738a64e79723c124bca450bfab9b0d5c77f422c4,980ba26d14ce4d945a1b2d419ea1dd1c4482a6749a3d3d31143d1bace113ccb3,2447bff548573484163c9372b0aac2b98d92e1c2e82d489b81a0258a1917f4a2,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,68.444,68.4,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
zai/glm-5-3,GLM 5.3,Z.ai,open,33,livecodebench,LiveCodeBench,coding,eec8df8475502a5e74ade0cb61fc4be5298d46568c47bc769366855d2077b8b8,2078ea9d05da867bd59b794f21b8c671b4e238422433b3702e05ffbf1f265c65,a01d5af3a2f30cd0722c5fc146d53c6c04aae12d22e883ef67a79c3717e7123e,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,80.535,80.5,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
zai/glm-5-3,GLM 5.3,Z.ai,open,33,scicode,SciCode,coding,c2dabc9866e165521cb71d99f10bcf017f0229515504a56ddb7c6728aa4854aa,7328e4104ce853178cbbd5ba877e138bd663cf8c29df46a156c3c23828a8ee95,89cd9987488d8fd48dfb9de446aa6009c6f3091d68659152bd782ad40846aacd,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,59.0277777777778,59.0,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
zai/glm-5-3,GLM 5.3,Z.ai,open,33,ale-bench,ALE-Bench,coding,5c29266c84f144d9cdfafc9f4ad40d400de72a1872b1bd4412aa6b98e00a046c,1333c4df07e57cf1551407e7297e3fd1656b889138a1e3d321fdd15a8caee27e,daa02bfb234208aad8d174c45bd774f76c33f7031a76c9bcbe564a266bbe1a38,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1317.4,1317.40,exact,canonical-source,profile_7b670530e689719e8ea1db1f,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
zai/glm-5-3,GLM 5.3,Z.ai,open,33,vibe-code-bench,Vibe Code Bench,coding-agent,f55545695b1979763f97834059776e7cad7008613cde2338da9c65c3a4244d34,4dde267f9a802865a7e16a706711151cd5ba003671d42ab1085e8932820ceb47,b6051c8101d2b7cc889b1734d5a4e75ad0b8206a7288c8627e83ede943c651ad,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,78.125,78.1,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
zai/glm-5-3,GLM 5.3,Z.ai,open,33,webdev-arena,Text Arena · Coding,coding,6d9ee69164e80f1b6f26fe899ceb18d4b992620067480c35952d3166f000344e,59f01106f5520fc11924d747f60bd707833cb6dd53e952864b40d21f35726c48,13abf23810655d5e3cc7b26658994dc532adf26433b34ce6b3c9855fa19f8fb5,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1622.8098625429598,1623,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
zai/glm-5-3,GLM 5.3,Z.ai,open,33,programbench,ProgramBench,coding-agent,b66db489774abd287059ad877e94c87fc2c8d68fad473b7d910babddb31f355f,9912b750f29c0f766248e153309080c51a10e05f6ed167bdcf8ef27289eae92b,e162a0017a7d5c30b94f5cb3ee6bbc8451218852cf10d356b289c7f62137fbe3,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,1.5,1.5,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
zai/glm-5-3,GLM 5.3,Z.ai,open,33,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,fddf942d9f6830991dae346f0db9fb66f261a4cde7d52526650b2f5e77ee8421,dddd8687bfa8467ad67febc373c29e816cc725107cb7fe3170bfdfb24f86bc0f,30e9b8b2223ee26132fabdefae1541d2cca015421958824bcb0a6c5680a032aa,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,71.536,71.5,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
zai/glm-5-3,GLM 5.3,Z.ai,open,33,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,17bbc4965095b6dc2b207d6d51780f647026ca0d217e9c0d566be1ca5b79e9dd,5b368289ee82a1e9a670c6fc9a669e98010b8e6f49222fb4d1febef9b17231ff,97de83d802da76ea6223c39ba8381a8716d109adc14386525e270ddf21733c9b,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,38.889,38.9,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
zai/glm-5-3,GLM 5.3,Z.ai,open,33,sre-bench,SRE Bench,terminal-agent,5bb5a348dbff33d0e8e03798020fc396bee7524ef333db68b12e3a20394bbf2c,a8f7c908cfef0c15928cbae00cda920a144c55b51e570b7b38eef32aa4a573c6,7692577614d8ead3402d7a1ae7afd0bbc29fbbdf8509a2d8b9522dad9168191a,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,9.16,9.2,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
zai/glm-5-3,GLM 5.3,Z.ai,open,33,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,6925f1aafca0596ee8d47a59130bed9b8a2f7ad57df6b613bf15dc06ad55bc33,5ff878291820ec03902381d41fbab24da0b887194277fcf3e4e1a13f4b2b7402,79804289c5074fc1a679d850742a50336124267abda7f3f0156dc998121ee294,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,50.309278,50.3,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
zai/glm-5-3,GLM 5.3,Z.ai,open,33,mcp-atlas,MCP-Atlas · Scale,tool-use,4be6cfe32884449de35cd8be816ded50da1189735a4d3161218647fe0a78eaa7,c36d90d0c7c956eb12d11465fa03b7549f1da625b14c008307c9b5d5948e202b,25c7bd980129449a1f4ca36872ed42b41f2faaf44475843821556178cafadf18,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,84.2,84.2,exact,canonical-source,profile_c742d73fd887bd38522e03c0,reported,unknown,,unknown,,false,2026-09-17T14:59:54Z,fresh,44,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
zai/glm-5-3,GLM 5.3,Z.ai,open,33,apex-agents,APEX-Agents,tool-use,adc8c43cab1edab5240e75bed2bb554b90cdf99e6b4ec09db28d541da7550043,d828f11900a5d001a57e8d7e2bc3b0ea023479d2083b1a8edaee87918b10c5b3,1be876096f3e5feefc445a69de9145dc897e11e1c7be0ca8ff733629b5b28915,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,56.60000000000001,56.6,exact,canonical-source,profile_628c201ab8499b38d223a8c1,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
zai/glm-5-3,GLM 5.3,Z.ai,open,33,arena-agent,LMArena Agent Arena,multi-turn-agent,ddb481236ad71f744b55eed285c0285bbebd28f97ccaecc3edfd8ffc700e1654,38a6a72d4f9e0cd1f6d6b38c0fb75ab7e0e5b272d4df548409849629aaf2679c,a527c9953e6c80676206346d2808a609b0af34c21ebdcf4143dde7fc98008a2c,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.009119361444542019,0.0091,exact,source-specific,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
zai/glm-5-3,GLM 5.3,Z.ai,open,33,vending-bench-2,Vending-Bench 2,multi-turn-agent,35dbb601e6b7a87e1e3e9b925e0078e8c20d0b51e2eedfdb6150c9c8ed31f15b,71bc0f0c84ad34051ce3451aab03aab759f6754eee83ab780d2647a7f58eda1d,460f57bc4832717e13967632b0bcbe5ca1bb1a7e4fe87aca0ed24ae8a70042dd,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,8163.609999999998,8163.61,exact,canonical-source,profile_c07ed0ced91381c8fb6c9cb5,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
zai/glm-5-3,GLM 5.3,Z.ai,open,33,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,3311f2d27627581d8ba418eef391fbb09a8c36c716b34b066e1d6dc98edb9694,3ed9e895470f302314638072fae00b86d07aeccaf57842afbb045bfb21aa2636,29d3145b34e4316d5ca415e192139ac1c8d29c637c6ff46799a03fb9557bdb01,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1656.88,1657,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
zai/glm-5-3,GLM 5.3,Z.ai,open,33,terminal-bench-3,Terminal-Bench 3.0,terminal-agent,937f41faeead6c07f404fe8c0532b34f770fce9974408fd6e8f5ea900c1ad89e,bbad5dac0df5b973c869634b5dfe9cc7bd9a7c39e5a03b4408ca0fea978a19ed,53ff18eba7ace83101286cb02f5020ab9cc5293ebd9dda6d00301c775e303b7d,release_ceb36f25882d8cea2c9d084a,3.0 · b936386b-4afd-4ad7-aa7f-6b52bcab9853,official-74-task-leaderboard,accuracy,percent,true,32.43,32.43,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-08-19T03:41:59.486409+00:00,fresh,73,0,A+,0,A+,0,A+,official-leaderboard,4,official-leaderboard,terminal-bench-3,Terminal-Bench 3.0,https://ofhuhcpkvzjlejydnvyd.supabase.co/functions/v1/leaderboard-read
zai/glm-5-3,GLM 5.3,Z.ai,open,33,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,9bd78720050a688ab3f317250fa1296f3841fb7363f8098f090314a37316f088,82fe1099b9358d53845ae53a6ec3d9d6f8c5f7c1275f3e11145a6089d85df308,7cdcabc4244581b79d8be123df4484760fec88ad3680add0dffc3d4e913b5e6d,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,5.714,5.7,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
zai/glm-5-3,GLM 5.3,Z.ai,open,33,weirdml,WeirdML,coding,9db3fad1fed400588c3950ae20f8f82d44892ae7331f0659c3c4247279d61104,3159044362855ea398e862994e6b4899d7bab74b9479119ad2f72a842f5ca8a9,d235c007ddac6805eeaf070ba3e06dcefb7f1a6c7cd41114fa61624a508e20ee,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,75.4,75.4,exact,canonical-source,profile_c9ddbc6765550e54adee760a,max,max,true,fixed_effort,,false,2026-09-02T21:27:22Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,epoch-eci,Epoch Capabilities Index,capability-index,956e99fbab1631277d72993a8d386994df0326cc9d6766139ac96536f0a6a45e,9b0343a1d13e639c842aa6710c93e108c4b521d02eea90d0244bb50b3e8e68c6,cae1e130187cd832a913603345d4e2fbc19053508bafea9b0e8f2be12f62ca1f,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,156.11,156.11,exact,canonical-source,profile_dd74f317b70c8284099eda37,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,livebench,LiveBench,general,3aef3fb0a6b34ce87f95b83427a254bb1b6b6d83111521a8e027e6dcea7af6f0,9d5645fa04e161ab14280912d8ea3097070ab3f71be53c454fb7cd115bf70013,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,81.10970238095238,81.11,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-09-10T16:33:51Z,fresh,51,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,b94f49b2a7f17debd460c998ede46067b772c5f476b47c8cdcb4915a2c205ba9,9eef60794805d1e4afad57748ded34bb7f91de9cb61e52085810b47469fc2e50,6947fd6a3208c0e07be0853cb9fc9857d705337e57801959fd2176d772ec06da,release_5f3023448c2ae4632fe1a43b,reported-runs,public-multimodal-mixed-run-configs,accuracy,percent,true,36.8,36.8,other-supplemental,supplemental-source,profile_deca33972debf2b7e8979c5c,reported,unknown,,unknown,,false,2026-09-10,fresh,52,2,B,2,B,1,A,evals-report-verified,0,evals-report-verified,evals-report-hle,evals.report · Humanity's Last Exam,https://evals.report/benchmarks/humanitys-last-exam?tab=scores
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,a549a28c45288e384af60d6acd36b2b0ec03c824e351cd129f9accfe0cdea597,37e1108d8972f1f9a2effe8cc0d75563d320a58f831c8d9f4aef2a54b4701290,d129e639daac4d14823fbdc31aa765502ca6c7fb4f60ae3f01bc0843aab699ff,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,39.249305,39.2,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,simpleqa-verified,SimpleQA Verified,factuality,10dbf662de063906360dd46187540834c04d6c3106d607e4e32303936e58c319,e6be99d0474c0c147e4d4390ebab0b8f933fe2f7dd2481ad20e1f8debebbd920,99513d1f10f76d7497af2c7f5b0858de5e9cc8911a52cfe316032976a6262784,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,42.9,42.9,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-13T12:59:33.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,gpqa-diamond,GPQA Diamond,science,86ff6c6fb8dbdc0a28fe15c2e30b8627db437ba04f35362690523574717bc335,6eed33a4c5178e52e9b74cccddf577af4ff1af0a1e13970ca67fa9a473976c84,1adeee9ed6a853cb95a5345df1eb9ce011de80431fb2509bfd2a536453dc1955,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,89.83585858585859,89.8,other-supplemental,supplemental-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-09T14:38:01.000Z,fresh,22,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,7cc1d851fa0cc4d82f8b3f1bb149f39ea4fbdf6f87c287237f36f019ed9e678f,1c17cd8db1414d00eca2c0b711214ef7417df4768f3a77490f6d8d1737a3b3e4,e4d403043222b5fcb211f56d28c7fc6fde0c88d3b6c4995890751feb21a95993,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,98.33333333333333,98.3,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-08T19:02:32.000Z,fresh,23,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,frontiermath,FrontierMath T1–3,mathematics,4f77f88934bb9d9b8a28aae4c3eb323aca6c67d1b38c90d7b010ea65a1198f19,24957f83f71d0124edd02f47802205f94466e43210e1353000f23717e760d8a5,e18f09d9e2f96c682a1cfc80f558b03b924e77b939246a910272437eb24d3db3,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,67.36842105263158,67.4,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-08T19:02:32.000Z,fresh,23,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,arc-agi-2,ARC-AGI-2,abstraction,bd9a0ec046188a9827c714ec8fa7ae7b174038f8b66ed72223696df9030dd29f,deb012c19f0574791f64c15dfd60d3ff2d7dcc8f3b7dfed55d44968ec1df68fc,02f739ab9bfde8e6cff4bb3c7ef004afae17217657b0bed25be6ad047596b341,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,72.91666666666666,73,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-08T22:23:19Z,fresh,23,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,mmmu-pro,MMMU-Pro,multimodal,51a0abe7d271e3faf042f8ae3964539493c8d9358c9e153fa996555e4fec01e5,aefce96f02d411abf2e3bb4071eb63f22756aefac712dbbb5202d640f58ad8ca,e6762ff11a884531add15b12eb46f33dfc18003b8c7fded2d356ee122ee97875,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,76.99422,77.0,other-supplemental,supplemental-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,critpt,CritPt,science,e543c958fe8850c3e18c91183fff2b7e98320f0ee5a5434a1cc621f38f66b90a,f4ccb2f5f9f5ee24b2aa3d8712776ef74ea4f9bf8c7f0539cd371459a9fbc67d,69d60d2e51b434f80742134cf07c7ccc7c730aa60c1a1a99ee9e359fd286bf39,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,14.285714285714299,14.3,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,chess-puzzles,Chess Puzzles,games,fc3a2cd3d5a336faf5c30e4833cbf99ffc251d608cff746efb5d5b2b7453cc2f,234fc9863156493915473bafb2110a104db7b827bc7c28dfa1037b9c6496c80d,693d34b93b9418c46ea84fbca1e42e5b0dbb9205a6f06091bfd93c9e4e081404,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,48.0,48.0,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-13T12:39:44.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,arc-agi-1,ARC-AGI-1,abstraction,241778aa5e55c9b9985ddf9120fb2e4ca56b35fd893f29ff20f6e0d5fd9a6d48,b5ede47fcf1ee0b57113ba2901ce7731f8f9914d7b7e237403e2f346a1dbacc3,a14ee82ca33c631994a647032530181e7f1805c4b03e97ee78f1b7cddb6284d7,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,94.5,94.5,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-14T07:48:14Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,swe-bench-live,SWE-bench-Live · Lite,coding-agent,d4d07a5fbe135c2cf7541346d88d7a7ed85b422c9806f4594270b49775490ec8,349e0d0b9755880c623e3bec3f1f360c830d9b135d9b3de3f5ffd6611eb8a338,d61dc76751ffc5acc0426b88635a8c235020986766903d39e275b088307347ba,release_9d2de9d5be7e9822af19f7b3,SWE-bench-Live Lite,lite-300-python,resolved_rate,percent,true,71.0,71.0,exact,canonical-source,profile_58c877d1022e326e6412c759,reported,unknown,,unknown,,false,2026-10-08,fresh,24,1,A,1,A,0,A+,official-submission,3,official-submission,swe-bench-live,SWE-bench-Live · Lite,https://swe-bench-live.github.io/reports-0605.jsonl
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,frontiercode,FrontierCode 1.1 · Main,coding-agent,1d1ac47017734aca650838a18b2e79990af4ee20925985e668f1bf12029b4d64,034e48a8c4370e29a2adb2966acd548d8816373ea2fff84c765b69a40192f900,bbf370ef0eece14387c38d3d6769503489888e11cceb06a23d6c8b3d2646f4e8,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,43.87,43.9,exact,canonical-source,profile_acbed0464c627a0050ab0f55,high,high,true,fixed_effort,,false,2026-10-28T21:25:01Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,ioi,IOI,coding,6e61f04b5343f5e28fa68340887bf5201a00c585b163eb32b02de9213cf8906e,1ed022570f854cf6370b2ab968c868c878d450e59f578012028d897e18855151,3612faa886780f7a5b684da0fdd128172c27241633efc8d2ece7654f9e8b64b3,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,40.278,40.3,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-21T19:24:33Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,scicode,SciCode,coding,32441a8a2989faf3097e6ecdfd4757996a3e79d595fbe54daaf5820c0ede35d8,84a47e4ed13a75446e3ed9c1757396dab4ecdb22defdc3e540faa0acc6d0836e,fc3067b1b4a2cbe74698f55ae61b0267526b3d6addbe1c9e8a12a15083350e59,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,51.851851851851805,51.9,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,ale-bench,ALE-Bench,coding,c0b4c99828fc1e1ee11a735496571e51bc24beee7f89260dea5cd3d949669f82,d3cacc3839e105bca03f2906e30a0f719c858ce16226669c1b07cdd7814a4023,72f06cbb6509a75a6886572ca0d438e3c26b92729b2fb1be0ff3c8e5a5e98e30,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1092.35,1092.35,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-09-30T18:53:16Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,vibe-code-bench,Vibe Code Bench,coding-agent,262a640547f54662053b26650ec58412a07cfdc5ddcb40f62f74139b76692785,28c6231437aac9df58d0e477bae25749053d82912eefa2a72de8d31fd6bd0fee,2d0fab9a64d382e3a3c1b2bd6bd3f275012b49c6d646ee4c1f8752923e8ea375,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,84.739,84.7,exact,canonical-source,profile_acbed0464c627a0050ab0f55,high,high,true,fixed_effort,,false,2026-09-10T20:32:05Z,fresh,51,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,webdev-arena,Text Arena · Coding,coding,c5aecb5ae1759574959e5dce34ccc7594a839310f23ddc7df08242b12f07d495,ce1fa28bf682b21e0960c92235e1e995b897871bc3f984e99b35c9645a7aa274,e80567596a8a54dc9ee62907c3a08b05306845c1e232cde8949d74c395f3f1bb,release_65fc140e7ffbb7bf62021f84,Text Arena Coding current pool,text-arena-coding-bradley-terry,bradley_terry_rating,points,true,1613.95,1613.95,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-webdev-arena,Epoch AI mirror · Text Arena (Coding),https://epoch.ai/benchmarks/webdev-arena
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,programbench,ProgramBench,coding-agent,f7c49200f910c5fa5b3e13a09b0bd4fb6a61d6738bedcd65ca56c5aa7aa94169,0f4905632d64d728b95e18b8b99cb43b3336f1a05d833668a0a6d4e8b95cc53b,b46813122497c19c9ba0dae80d148b5b82cc11678b843b9b7407ed1654c44eec,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.5,0.5,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,weirdml-v3,WeirdML v3,coding-agent,d770cec83a3e5a01cd6c9059301ac4c568bdd682e39eef3eda6e35a7864aeec0,b23e37889e728a303dcd822b958782f793b4bcab20681625b760da2f52e30b84,b6f1ea3da24ab139118d7c35bbb8abcca9cf0576d14c800f5bda33d0ee5d1431,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,5.946331818181818,5.9,exact,canonical-source,profile_acbed0464c627a0050ab0f55,high,high,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,efefb63c016ea894910bbbba459724a1e918f44424524273d5819064e61e9fcc,4f2914a7f8101c43a15e762f644b3825f0f673f2f3542bdb642c28128ed195dd,17d94bf92e11b19591bb3497ce2fb50b9bccbf5c8c309e60ee34d8b26b7d05a9,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,74.532,74.5,exact,canonical-source,profile_acbed0464c627a0050ab0f55,high,high,true,fixed_effort,,false,2026-09-10T20:32:05Z,fresh,51,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,1429d0674c3971103fba32a7554f4aff502dc596bf9f751a4ed2a29a4c5db9ec,7bc61c07eb68f4e4de010e1c9cb8afb5a262f6bc31805f72ac954c2dbb2c26ca,8c47f22bc49e6d3d3d773e8423926e2299a6e1b68551422ce89b9e2c4b531f4a,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,19.697,19.7,exact,canonical-source,profile_acbed0464c627a0050ab0f55,high,high,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,sre-bench,SRE Bench,terminal-agent,041d616818cb9cc834a6c52d9f5d11aec65b026e8cbc3a9ea0e356b1f9a88c6b,f8921c01d4c1b76db3445ac48c0903ada3ed1c0a3f8532c96b0d97311301e1a3,0919aa683212c45ad60fa265d765e025164b3b948457272c7b4e843719e2b126,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,0.763,0.8,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,apex-agents,APEX-Agents,tool-use,7fef241518a4c21e9e2f7bf96618e849ca49cf5e9b05ef30c1145f548ac7b3e1,63b35594053fc5513b04a6131ced003c54b9f438e2cf05b3bff6b874926f2782,d0b7f967aaa0dddb512a04b8a69652f86c5493b28ca74ee7a7863abec3a2be33,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,39.5,39.5,exact,canonical-source,profile_99c4153bd6fcfe58f2654e73,reported,unknown,,unknown,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,574d8e6f02133c61377c997200383b37061212e48feda17ac17b396e8dd7a411,17667f96630b20383ee0b1e962bf83a7df5ce7324948079604c56163ac719c8a,6486a9c65218c13ce168036d046284a375175225634d5a18887e72c5bc783175,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1600.0,1600,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-01T09:31:57Z,fresh,30,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
deepseek/deepseek-v4-1-flash,DeepSeek V4.1 Flash,DeepSeek,open,34,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,bc2032b691438f6034ce1e578310d49a15c1b60fd877de6df78d98e5141e5ee8,e604eb3299dd9533c7a82034767b69969cedb50e6581151307c7a7b908fd02b1,ed35a8b89fc4f0b309ffe34eba31aacaa3fd0f341bb3c6e2ba2359cf53927345,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,0.0,0.0,exact,canonical-source,profile_f5050027bca49136f8c18c5b,max,max,true,fixed_effort,,false,2026-10-26T22:05:28Z,fresh,5,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,epoch-eci,Epoch Capabilities Index,capability-index,bb21a3570bc1ead93e589f49172aab7c15a96a3812a6665cadc71fcaac1bde60,adf14665e033b84629b7a9a7a03f3ba8c9d6ddac659e045db7ee7d76ff5168ec,2cea4b5221c2504f9ce4b93d4834e13ec1d94416ebc998a6a3559bc13bd7c162,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,155.87,155.87,exact,canonical-source,profile_c09f67f64bed3296f4de97af,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,186d8a910440e294aa1efe57e6a199bb92344c3303f2cd1f4cb4e24c55fa3e58,71fb16f148c4fc7d947d2f908c7c1555694739ba807f179ed55c7edc1fbcbb63,737c5486836f0b46e0266462687d4eaa3851cf8d85e9fc717469bbaf7534d901,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,23.063298,23.1,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,arena-text,Arena Text,preference,544065ac25b698f04b2bb2c4f2a684a82fd084b79b5fc12327211fba18665611,8d2e164f1685fc88e8ffffc140826c6bde2395baf834dc295af4ece0b044bc01,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1465.0587343984691,1465,exact,canonical-source,profile_30663e63c085ce7e9e35d6eb,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,livebench,LiveBench,general,fd1bc8d80a551159ab1d00ffd59a7ac7d5f2967bd4aef7e2497bf4aff557b533,e4cdca75b0fa369cd54e8b605d97586029d61e0eaec499e4e2605db7e020e32d,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,77.43614285714285,77.44,exact,canonical-source,profile_2a6ad306680e2f1f01edb742,reported,unknown,,unknown,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,76e6fabeb085a834f0e7d3bac34867becf3e12b44dd61b68598b8985a5b9b128,7d9433fb208028a63a95bc36a97f4e7b0326f9b492979994c13e4b00ba7f4aa1,75c0bf5cc62a5e8a5b5579767a3b94349eb76845210b6368a4b0ed833f857d74,release_1f4bfa5486eb4deec4942449,HLE original,provider-report-no-tools,accuracy,percent,true,42.7,42.7,other-supplemental,supplemental-source,profile_a6b20b4d1f7b9f4ae3e96950,reported,unknown,,unknown,,false,2026-08-13,fresh,80,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,deepseek-v4-pro-card,DeepSeek · V4 Pro 0813 Model Card,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,bae6152e34bec1027bd585b40a5eaaf8ffbc4588b6089278404a0aa80863b05b,daaacad00c7a9c0d4442f6b22726c4824df42780ac7fd7ea9c8712f474518216,6bb682f08bba3be3731329c1420e9222e1c05ecd23476a8a152984bdf45cd149,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,41.010195,41.0,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,simpleqa-verified,SimpleQA Verified,factuality,0c70879a0182d77c9be377defbe1f4d77c98249483d831b45f2caa97df4e7e1b,fa3a22de174183bc711d3dcf1951777492729ec9ece957a1f53935baf0844421,fb389f5e6ce6839d1fb8f0a8eab180a25968c0d0c926ccf278768bcba9be7348,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,52.90581162324649,52.9,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-08-27T19:30:20.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,gpqa-diamond,GPQA Diamond,science,4a23194b844278df50c9bad9b3b57eb06ebfc51b53abfebb24c2a051b47b0d13,fd1c653ca14a23f547e2c7c71c957134b4eeeb9eeb0fe74c1dba0fe1a3b6c63a,38161a15d939b35f8ffd6361497209747f240559523d5748d1ff72396ca69bfd,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,92.424,92.42,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,mmlu-pro,MMLU-Pro,knowledge,974d62f1598801bec0dd3f7137f43c69979ac867e920c495d7b76c6b5da1ea80,15d6194840db2bf76241c4c6780d5ffba503ce63e51bc9e51cbb766823439cb6,667fd49cfa7b9fae806e1b03bbc1d0fe3aca2ea46596fe3411aa0dc2434b4e6c,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.972,87.0,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,ed8ac7e68728557cc99efb9b94a6d1ad80c46b1819786483c5a3ee38733e58e2,0dc66ceab358a798bfc2a6c85c3cecaa0a5e03b085a27d747c167459adb4414e,93edc8b8be7acc7253a111ad5e77411d3b0836ac0cb76c70fedc517ebc5c97cd,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,98.61111111111111,98.6,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-08-18T21:03:23.000Z,fresh,74,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,frontiermath,FrontierMath T1–3,mathematics,d6fe6fec63927e8421724d748d545b4d504ff7324300ba7bd45297aee007d8fb,3333583bf005f50d16e16a1d2b8b4c367cefd8e63f5461d6918abbd9e3966050,3a7dbc6bd2b185e89d1634f1e9b098cfb3a5c5a2eb6a502f5c17c352164289e9,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,64.56140350877193,64.6,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-08-19T13:35:48.000Z,fresh,73,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,arc-agi-2,ARC-AGI-2,abstraction,5acaf5dedecbabac9e57c1a87b06ca4a31eedcf7c1b9b41ef18195141108828c,c4f2c7f2ca59e4a402f05da988eef15d0f4c876bd68d4c01e21fdba9973cd611,f54664a069cd3a263e222b62884cb95777c40556c3db565b8f8b542e48ca7239,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,61.25000000000001,61,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,frontiermath-tier-4,FrontierMath T4,mathematics,d407ded599c962a96289bc0c809a99bf39192ec87f90802e188cf4514af01113,a77faf8564216f1fb09f425fe81e628502aa09591950edbb1eceecc7b3dc6b09,65a81bc4e77c19398496f8304279491bf76b327f284447513082d40883dded3c,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,26.82926829268293,26.8,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-08-19T13:35:48.000Z,fresh,73,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,critpt,CritPt,science,b012e13b97c5c297d19b0d6cfb31daad6e30c171dfd808de70cc70e86b6a67ab,3d31cc276a5fac45563f03a01df7e32fb0d476e61e9ca26359d87626995ae1d4,1abd067406f487e8eb4e47e9a741e294e5071feab4d4584b9ee9d25afb0439be,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,18.0,18.0,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-10-28T21:25:03Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,chess-puzzles,Chess Puzzles,games,d99c44e96715c7d554c3671cb9fea56446fafa84d530d20f1c6c08f2537ec094,ab46dc19612aab455ce0c7cb209f54e23be904eedf7b3503d1212115c5151606,176acecf7dc6a84d63ccea8c01beca6ebd05693417c33427592398e8d7a4f9da,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,46.0,46.0,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-10-13T13:13:34.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,arc-agi-1,ARC-AGI-1,abstraction,4b6cb7e8845c02bf694e11ebdd649458f84ca6965f4d0c814b73c02a3a75730c,e23a1a43c07df7541144a69d686b6fdea9205616a4c592d4157c763a4eaeb9ee,20ec2e9d0c96a3ad110511dd3bc6664dc17452fcbd0bf5682947eab4b984831e,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,90.0,90.0,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,swe-bench-verified,SWE-bench Verified,coding-agent,2e5b739d72b06c0a021aaa5b6d109c7d736d017e1ac3b0019798136eb231d598,6bc81882f409dc7e0ead091571bfec07a5f558091f797d9d4c4e892ce5d02ff6,85959588eff6b49765017779c33994a2669b391ff7c12b230bd90096006b14c2,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,96.4,96.4,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,frontiercode,FrontierCode 1.1 · Main,coding-agent,d788573b6d127df4c3bbcfd14cd664ae546f7212e000712d50bf8a66f88cebb7,a63ddb8811e3b99aa56581a47bc96fb95ce41e3ef677a7c10a98be74cf76d066,cdfa9576b0fecd1b9913f0af0b92aaff679bc49f2dfe1fe6daec01ce5745a612,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,28.549999999999997,28.5,exact,canonical-source,profile_30663e63c085ce7e9e35d6eb,high,high,true,fixed_effort,,false,2026-10-28T21:25:01Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,ioi,IOI,coding,970c70bed6dc79589a5d79838d2815284a1bf63e9f009630909d722492c8e79b,5da6c552b96d5e53df05abe712fd5d8a857287ad4cf16d96d4fb4457c6402a35,8c9fd853606ff405b8afab9da85a41ad682f0b0d2c881df799a68b48591de76a,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,51.611,51.6,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-09-14T23:34:42Z,fresh,47,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,livecodebench,LiveCodeBench,coding,2325dc64cafd923b73cba8c82be9691b7b7c54c29b2a35d5403a8353e99df28b,2204a70458d671c36bf3fd8503ad1f2299c87ff2eca3a44819aacab9fc9dbd7e,2de02e83a78a2f7ed2387c95d0e5bc5447a0e06484a3dab91875f46758d1bc8b,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,87.525,87.5,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,scicode,SciCode,coding,ce9d58f3b8daf5b7564891ad5a0c5789d4d30aa1c415320b65ce7d52c0623cb7,128638124d99fbd77a832487ef4a19e0ed8d91c21c17addf6d7265aa996f2da0,b57fc33a04e0622780b6690f5f931d7db1ed22ca076316648d4b93926e08bcec,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,51.0416666666667,51.0,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-10-28T21:25:02Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,ale-bench,ALE-Bench,coding,0619c70fda5fee4e5e40cb85ebe91c75674c2bd1b5dc4e5c0c2c4b9a2a495a8b,25547f3ec5d8a5a4ecc3551d02fc1fc52aa0ba6f3793a27355de40bf329c4283,6891d233e4e51d5e32e85f40613aa84ec70db2d12f8707e09e131f6272670e8f,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1403.15,1403.15,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,vibe-code-bench,Vibe Code Bench,coding-agent,f367edabdb25ee12fe576d5a71e20c9c00bce88fcede81cccbc14fbbf814d3b7,47e6319fa3a363e9cd194b62d0f0a7c6e3df2442c4ad74d671af6cb915ff781f,db1cbf840bb13d5c6ab99abc75b9dd09c3cb9679a3b4072d9e450952478dc5ab,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,82.297,82.3,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,webdev-arena,Text Arena · Coding,coding,b2261b975869227f8fb4c68014b6472b07103433a5196c318cd8e5341b955497,06cf7c8bd6cac4342afa9dfe0da12404efcd0dd42b7dd3e08ff11c3ca6f9e114,659dbdd801269e5d62a98df7445fe89c67eb9994266484530e0657a7dc14b8e7,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1581.8580682830898,1582,exact,canonical-source,profile_30663e63c085ce7e9e35d6eb,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,mirrorcode,MirrorCode,coding-agent,bedb9b1439198da465c9548b35d708026cf3c0d300919179c267833fba2a5a09,17a2ac60fd66c13a1385c45091b31a1f5620214c38b8c643fa7565f11063ab72,38173aaa8e148ac5065daef9cf9c11c853a0cd666b92b00c3e623b75029d1593,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,33.33333333333333,33.3,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-10-20T20:32:20.000Z,fresh,11,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,programbench,ProgramBench,coding-agent,bb26c2da83b85d692942bac39c38d054e1c551fd5a0e1228130e724eea208317,e841b993f205c226e2be71ff63ff1594035838c3a1c24385c7e68269b3ae790e,686d1b2df13b91df77b1f42792ef61142ba1d97d2ffc5364815af261fb173003,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,cb1b82b597079da65f349b1efeb07131f74e0f4bdffcc01f982d29006c122eea,ec607d0c57f259043e8784e42044a4986a1f9cbcf1dc3e2d42059e254c576048,a881c122281ac49a833fae875e28b55192135f36775b5330f25137b63ccc61ae,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,54.682,54.7,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,d5f6f9d921fec354466ca278037924c6cd3eac65486b1f9f7947e4718b0a2384,17054dacf3122f236eda11ef9500738fa8c9db37b426b8d0157fd79ab2c7812e,3bf60d629e3747ae7c95c4dbd78e331c7bf6b91085409e34268c32ef1118f485,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,14.141,14.1,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,sre-bench,SRE Bench,terminal-agent,52e3e7e8b4408ad3e68fb08e28f74aead797fea4416b9d004caab3ca14ec3c3c,96f043ba94ab55482fff924445059c09edc6ffff0a79a8525c1a2bbe9fd2089d,dcada9b2667ed80333de23270d9b430a382bb0e7b5a5993c3f0ccbb0f3bcc14f,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,1.908,1.9,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,f08219f0723920ab4c1c0d0cbe2309cc508c7fc755083ce41930e4adc34dedcb,3e40072924433de7f3e7679956cdba839dd213e8409b836c2bbf118e37f143a0,b99bb0b1f3613089c782db885bcc112434c645a71c97d64630c00d210360f35a,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,39.587629,39.6,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,apex-agents,APEX-Agents,tool-use,9e80a9baa97441692cd136721521753e0f1e81a3c622028c4a1ba4dba6758cfc,a3f38c9cecad9a00eb9095ac0fe5a2147572730c5a13bffc1558330e5fcfc5b4,b65e623417317e3d7f2f62acc87f19ed0bbbce8bdf0102ed0ee1a5ab8161ffe9,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,47.3,47.3,exact,canonical-source,profile_7bbfeeac30a4f4702a8667b3,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,arena-agent,LMArena Agent Arena,multi-turn-agent,e3c64f7a183b5fa60a2e2cee6e7f2d7d7bec1a0584bb23a96e7c70f3e692df0c,345410cad5ee160844eaee63b9108694cb04400ccd5404157d309118b001eba5,2c15a6f4527dea4db1569270c131a5d4bed094cbca7a2714355f863ed0ecf7ca,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.0030051721572174742,'-0.0030,exact,source-specific,profile_30663e63c085ce7e9e35d6eb,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,8505daed65f03092ed2df5fa52e92775dc362482835c4b1d8aa1832c2dd9e503,d96cabcd69e3995d7a4e1aa2093f38d71ac9799ed0758332a90b9a3db1034d45,b46e0dfcff8c8b9505f0ff101cf3aee62cd533d0de417752c9d19c03a18728c0,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1457.37,1457,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,a8dadf58ca154cf4806d2f65f3572d0043df768c325949258458a448a01e89f9,fda454b343ac3f0653636ad4a30ba852e1d601a4e5403e94b6a307723c1903e0,3cc9acd33160fca4968fd5277ec665057d116b030a62d0ba9d546262851bb0ab,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,4.286,4.3,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
deepseek/deepseek-v4-pro-20260813,DeepSeek V4 Pro (2026-08-13),DeepSeek,open,35,weirdml,WeirdML,coding,6800256c7e1221811e6662916dcb4708c646cfffc4c58a9466bb92a03759485b,5e9bcd6032c728b9e7293568cc533e127dedbc7500fb43192cb971856895830e,876d95f59e5e7ca1a61549d0c9418536be0db88d61f34ea6708f45f226b33e68,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,66.22,66.2,exact,canonical-source,profile_85d13172a494484884c0ddf3,max,max,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,epoch-eci,Epoch Capabilities Index,capability-index,5eb0d20b1f3e75aaa9b23b089af8f97b9475204e6b0cd875bd8b254627bc5216,f15ab99b05135be210af3dfc2a08631f68d4831e79571e801a7d9210e26021a6,1f26114cc61f6a9d17a72ada6f47a5f63bd79361f49bacfea508045dc0e3cd86,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,155.85,155.85,exact,canonical-source,profile_152e3a035044e671415749f7,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,arena-text,Arena Text,preference,ba55d207005e628c8bf813da7ddf50f61b1b5eb0d1735f127553ddd13782c22d,bbe1db75549247e203de3ff4dadf7cda18076d88dbc4006358f06015c0dd17c6,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1503.162754839902,1503,exact,canonical-source,profile_f3a928c2dc4d6711d402db03,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,livebench,LiveBench,general,2e42aa46e2aedb0d33feb9e63f24bd3f7bb3d1a11926d0af97b8152dbda97db9,62e8d2741367f013b032bb76d09e4e8eead22facd06f5934a8764451be2d0492,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,74.51990476190477,74.52,exact,canonical-source,profile_f3a928c2dc4d6711d402db03,high,high,true,fixed_effort,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,4f1509b2386d3f7c73d08032397c58a8bf1addc46150cd6ae83cfbee3e3b1d59,f1af45c4b3c2371917af44ec7a9305e21fd72f93537635345fb475d5bccfc08d,3f0ea962416db561c471a5f68bc645c18e42010eb2127df659eb6c8a0d77b31f,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,34.44,34.44,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-02-17T17:04:32Z,stale,256,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,simpleqa-verified,SimpleQA Verified,factuality,f4e0f501cd6c05f492c23dbac763bb77ec418ad56b01efe0d8018dec969834a3,5ffcf89edc0616fcd6d144f5ef9a4de0497bff501781b19fa95220e34213017a,c0cc28b2bc35c4d4af2d8b42dc97810da93d85493654fee2989049a865502419,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,47.0,47.0,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-08-27T19:27:45.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,gpqa-diamond,GPQA Diamond,science,c84f794ab12e986775ddd665bb6a6140ecbd4c56b2aceeb39c2e9e39e1cf2670,2fada5ac0388a322efebab9516838b06d34fb5c4169515659046960c66e39236,692d3831b8cbb7a644ec31cecda17e81066e6305d608eef2c408c62c850577b5,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,89.646,89.65,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,mmlu-pro,MMLU-Pro,knowledge,112cbe064997117d2ad46121570ded6a85392550ef1b6789ab5c6c2d7f044cf4,a4e5c54d65fea880f24ac2366d938cc7ef571c957f8dd152da189494c4418e4c,d2af2ea7d7ba56574b2ca9a9aee70a7ce41c605007b2dfa0635495b0a29b65f6,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,89.107,89.1,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,68417e1301fdffe4843e85d812f9572af92c298d2cecf71862864ed0541c65fa,7e94c925744990e140dbb90996586f902bed14892ed290763d4507f7d4901ad4,91f1ee70583fe8a71c9a36a257a8ff7af9a54946c9765146dbd06d21cc083bd5,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,91.11111111111111,91.1,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-08-06T23:58:17.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,frontiermath,FrontierMath T1–3,mathematics,55f484e95424c5953d21537aceb8afc1a7c382fd37d7797e06d6bdfc1f2a1110,8a3330741a609c7e0734eda278dde595e464255406e2274535287c8f28fc9dfb,4cfb756cf4e21c80a11a6c1a5fdd6d61e065e2c01ee1e7178cf2bc90817cd4fb,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,65.96491228070175,66.0,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-06-11T00:19:44.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,arc-agi-2,ARC-AGI-2,abstraction,b461a5b3e95ce185cb3217eb325db73b09eb0e39ff7db2e2b7d6242f74891a84,a21081add0f41d64efd61cff8bf2fcfefa52a37ca435e84a2dde82b2a80d9554,064163ac017b920b561ddfc144ba380b6232580bcc756f1dc732224e4f3c6eec,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,68.75,69,exact,canonical-source,profile_3b724677f8ab5619a6947425,max,max,true,token_budget,120000,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,frontiermath-tier-4,FrontierMath T4,mathematics,354e9e9e3e5ccbdf0f9f525667e4e6da9df671e8a87dda8e5105a399fbf2bc8d,c14a6cfdb2cba541e96ae4da05d4f6269ea07e2fadacd6202d311f1d64694769,977303192a8bf64f180a1f4395dc194dde5d277227ef9d8d4f9f292afbb127d7,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,26.82926829268293,26.8,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-06-11T00:19:44.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,mmmu-pro,MMMU-Pro,multimodal,b05f26e6bf95b4089e90b42c454f527750ee2ec985d23ebce701f523a339f799,e56a6fc6038d840f1de306739064f55ba30b034313c30e9fe32c35b18e31a312,751ee82cb51ee0248b91981c3af5aca737766de94a5a828f99dd47396bb6aad5,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,83.873,83.9,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,simplebench,SimpleBench,reasoning,2f0ab562852c86193e84890c85eb93b862fcc69011cec9a25dd8603a74fd3490,821fe1731435e8bfa0fa6c03baa41144e78e1a2e1026d7ba5417532422c58723,e4efa10f59195bbd55fde3aa51b8c9db51ed225321e6620142291c31b330a1de,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,67.6,67.6,exact,canonical-source,profile_774e289c6cfcaa16db364671,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,chess-puzzles,Chess Puzzles,games,5eca1223819a2fb09ee5f8c2db0cc3f00b489fb8143a8d54e75df6e4765a58f8,020e5f208dca362cfb6c9223730cd2d22cc07bd9a2d0b8836e5f75bf4cb7bc62,4d2dfab0a746311072c5b8ffbfc0af3a6a326784e207115c4d6757cbcacb759f,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,21.0,21.0,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-10-13T13:43:19.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,swe-bench-verified,SWE-bench Verified,coding-agent,c2e521d67e1b18fab054d408ee368ef7c9841098cb47f68d0c9bf46d489e8d4b,913c81b1b1d9ff7f932011d4e76095e282c53f8bb9748ceb4cb7065e7afade4e,0efba33b346c02c6c4c1dab6f884af894585e223cd00e12df247eeca85491f67,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,78.2,78.2,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,swe-rebench-v2,SWE-rebench v2,coding-agent,a3257c378b89451b6ad59b3698f23e7943fd81ba6d7a4fb0ef6da9ac1e9a657d,77bac46b049f106a32671c6dde0f7a1af34d593c009617f9a3b2a4ab9c5211b7,ee4e41d217d898acdf8d142ff31ef7c20f7cb11cad469979ca052b4bc43818e1,release_72d764965dd7baf34f983249,rolling-v2,2026-03-01/2026-05-15,resolved_rate,percent,true,47.81818181818182,47.82,other-supplemental,supplemental-prior-window,profile_f3a928c2dc4d6711d402db03,high,high,true,fixed_effort,,false,2026-05-15T00:00:00Z,fresh,170,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,swe-bench-live,SWE-bench-Live · Lite,coding-agent,e95322883c11ec76bd0064b1a02660f74f5ac959301a763eef933a3b620002d0,05c57601ba4663667c530672f544ab5a09d7a6e9d585f7681e01f6b5f977919a,316ad8f92bb102c364f20b335f7b61c5fa87ed7e9712c42d2dc4373709d36dba,release_9d2de9d5be7e9822af19f7b3,SWE-bench-Live Lite,lite-300-python,resolved_rate,percent,true,25.0,25.0,exact,canonical-source,profile_7258dfad95526e807989ecf7,reported,unknown,,unknown,,false,2026-08-17,fresh,76,1,A,1,A,0,A+,official-submission,3,official-submission,swe-bench-live,SWE-bench-Live · Lite,https://swe-bench-live.github.io/reports-0605.jsonl
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,gso,GSO,coding-agent,511085970eac586b96be670bbd6c89883f5e3fde9f05c6fbdadb588b94c730df,610eeba4e21515274874837fac8f618346b03dca5be4daa186373a0e7ab7ce62,1596cd4ede296ae563e39dff0e68f7793750895e0d1eac5471fb9e59bf002fab,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,41.199999999999996,41.2,exact,canonical-source,profile_f3a928c2dc4d6711d402db03,high,high,true,fixed_effort,,false,2026-04-27,stale,188,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,frontiercode,FrontierCode 1.1 · Main,coding-agent,5c6df73441f42f4c70d9af7f03eb58fb0860c15a314f5e247f288f57102210f3,3fd204224734556e81e1376d9ac7db4313c95d8a8d84ba8bbb181724b9756171,c748fa81cfe10c6144fa6f7d7562489be23172b94784012fabe2a240a05be6f7,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,26.640000000000004,26.6,exact,canonical-source,profile_f3a928c2dc4d6711d402db03,high,high,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,livecodebench,LiveCodeBench,coding,2c78c9a8bffe201d250b1756c5706ab29aa7cbda147577dccde0a88dd39b301a,82e0ff85a7a69b99775b31448cbc58e2855b41b4ff3e57626728d05f396b1a42,312178ee84b097dc528d62c68741bc228deb676ae4d0357f763789536981d7c0,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,84.676,84.7,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,ale-bench,ALE-Bench,coding,46e15d348a1e2e852d312f5e267a65d339a3a1fa3f37cbad2111e10758670b1a,8b0417ee4d06d2e59f37f0ed31312231a1b72af6c31d5c800b88a6e822270492,bd3d060a508306ef4cc3f8ba4bfb9d72658f96c96a6845870b108e54d6fd01d5,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,996.5,996.50,exact,canonical-source,profile_505fa327bb2ef52374adba4e,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,vibe-code-bench,Vibe Code Bench,coding-agent,c22b356f78040bbebb58a8c55241420632a5b0f8346a506d8ba8a6908059deb9,3bbc79fef96e2c06c2405b1c972b6a340ffbb162ba11e3aa441759cb8fd3aadd,a4a97dfdf352ec9a16a9036e7aa41ec9893577a2c26bda945a4004262dcd0cc4,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,57.573,57.6,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,webdev-arena,Text Arena · Coding,coding,f6a911e05b56f3185c3c69e6a7fe0075217fea5e37d465d7737adaf03a209d21,0288dfd38271275a4501f995958711578f44ec13342408b0554241027528cad0,110b4ca75c475241e184c9a2d8d7a99d0b75a64eb6f6cc83304e51d313c0a814,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1546.235576890829,1546,exact,canonical-source,profile_f3a928c2dc4d6711d402db03,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,mirrorcode,MirrorCode,coding-agent,615c1652ceb31e31e87bbeef916ba86701edc2c62c1ebb769850eb95ea2f32fe,1e0fc694e5dd9d70cecb310f81564170782847dfd8a9488493c8405cedd5f28d,6104ec04b8815f57ed721dc6946068bd2e23f650c78eddd6f09b4e3e97f6d68a,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,33.33333333333333,33.3,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-10-21T16:48:25.000Z,fresh,10,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,mcp-atlas,MCP-Atlas · Scale,tool-use,8847b9a1d434b0696a59466b3e5341e64b0fbc813a7bff6035e06ffa318ab468,aafc6e7c2ce5beeb3de616ba8eed0e34e2be8d69a79106fa990f33fa278ffa6c,ea5fc610230b9fb2086c43104e36e9bdf39d223bb6bdb1e4d10b4a87ce5b18ec,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,76.8,76.8,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2025-12-18T18:01:17Z,stale,317,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,apex-agents,APEX-Agents,tool-use,99d679f2f1468f0d172a31d7db1776821902cc47c55d4b2fa26f691628b7caef,0c948b40e9e83fe3b4c893fe0fb8747600d10661aa77394627a5154057d1f691,87c7d6d307e6af6aaec53bd62b0bfa5824dae9bd7f22f32a5d6934e4bcc324cb,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,46.3,46.3,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,appworld,AppWorld,tool-use,1cc1db4199d03d47da0120c97823e9d463f446b46557bf3b7f7ff8ed2929fac5,6c665940b7aa160398a5b32af2a17e08b0123d7e0dbddc617029489a17385a15,72b8a94a56270e5c3983fc0a16989987c068c2c329242dd70533a415bc929b59,release_3fa145932216c0afcdd9dae6,official-c1f56015cf7c,test-challenge-all,task_goal_completion,percent,true,72.4,72.4,exact,source-specific,profile_e22277a9c302417e104ce78e,reported,unknown,,unknown,,false,2026-03-23,stale,223,0,A+,0,A+,0,A+,official-leaderboard,0,official-leaderboard,appworld,AppWorld official leaderboard,https://raw.githubusercontent.com/StonyBrookNLP/appworld-leaderboard/c1f56015cf7c3441ff1933f5bfbec879798d7bbe/experiments/outputs/_leaderboard.json
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,vending-bench-2,Vending-Bench 2,multi-turn-agent,a4fd702a4daddadbda7b0557978393d5ce348c9beddc3343c4d4d740ed6ba281,87f2c612d65674f06f2cc946c7e71e95c2769ce2aae37c8c6b641f78c2c7d58b,5984834f42e164be2902d2931446c87ab2d68e88835f3ba18c49a083fba222b2,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,8017.594,8017.59,exact,canonical-source,profile_1f201fa792617714eed8a8a3,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,c03ae079b1ea3586971e1d618d87ecae0a76dcf8c6de94677ed9a4ba8710997e,260d06a9dee49bb1cda333f55abd10aed0b1138f906a2afee8c500498b49fce6,5a2403363a0c9feff2c6ac7d53571698b24966266565d4704bcda085179eb000,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,27.32,27.3,exact,canonical-source,profile_850fc6175572e341e7de5b0a,max,max,true,fixed_effort,,false,2026-05-06,fresh,179,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
anthropic/claude-opus-4-6,Claude Opus 4.6,Anthropic,closed,36,weirdml,WeirdML,coding,f9fea0954ce2f71ebd697d9e006556ac67b12d4b53ee7189c574ad6fb2815c0c,3d3a1e1e6ffec140ec9224c823e07d03446e5c57b70da45d60a7a596c1547301,e220d5173b7f9c3f85bf1d864cf45f1646f627d8293a8c535d4b9696d26b25d4,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,77.95,78.0,exact,canonical-source,profile_f3a928c2dc4d6711d402db03,high,high,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,epoch-eci,Epoch Capabilities Index,capability-index,13dfbb5bdac1e7cc20479d4832ff0dabf8adc8b432f0471b2427783fbd445951,c0f94175ade1a1db7102036e77207619a77247248ba281f0e7183cf51aec275b,b62aa23688286d42500a61eabe9d164440492fa42b5ab71a5c184744ec6decbd,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,155.24,155.24,exact,canonical-source,profile_77bb122ce308b85a06f0e2bd,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,arena-text,Arena Text,preference,6865f561b2469c25c63d3509323dc1493d51534c7ec01b2b25da03561a239e38,a5812a0fb4c9fa1077779624ac8fc0d13b94f55d673c7e177f547d920c488e4f,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1486.8999611035829,1487,exact,canonical-source,profile_fb779cfb20c726fd0ba130d3,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,livebench,LiveBench,general,ef02afb260f08014c0cdd896d1cd6d6af00ff63e878ec2435d18303040be0b96,bda3a61451750480c756614c6aeb0f5488f02cf7dfb5267c5043a54fc4c7744c,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,76.95184523809523,76.95,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,46bf5ce73b4fb21a8a4822833c120f364ad08dd657e5068c74a9dfd70771bc59,5af7bef9cd1129d37b2ffee1e8f58e9d9776d4a3232487a5c77b5c3292036722,cec1196e0bb0f36d656a32a8e6313a188a19f65c1eb7067d97142249ce305032,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,46.44,46.44,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-04-10T15:51:06Z,stale,204,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,2f7d80832cc3c4f491dfb14dc8b778153952ef475da81d5fb4296a0c7f842d8b,a0d9e721394552aada91bff32714f5d04f46c68c51b297ee9eba619bd800c767,c2ccdcea67309602be1cf557cd4582a5379241063ff487f6c4acd6a9afe5f745,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,47.034291,47.0,exact,canonical-source,profile_8bf7a013a6ba02a14d73d95c,reported,unknown,,unknown,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,simpleqa-verified,SimpleQA Verified,factuality,dfac9feb4d38caa47936a08f1190b23baa02652c83f780bacebcb7b52b9f0554,93dd795d91b3d633faab06b4aaf2ae201fb20f98f050482a8190b34524bc945f,6ec618647aae19d98a2e6a010349940ff6e286be1c872a82002ffffe0cac6a6b,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,73.5,73.5,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-08-10T21:35:39.000Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,gpqa-diamond,GPQA Diamond,science,c7de7f95a40f5d4ae7c59a6ec02b176e25af64185d451ff86e7b76ea472cfae4,ec8a376b44833c24503fe6ff5e0ac258a7263a1b62dcde6ff8509ecdb9f6dbc1,b61d787ce49587fefd9a3b8d192062edcee44f19c95db74b4d064a90f17152e9,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,95.454,95.45,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,mmlu-pro,MMLU-Pro,knowledge,71c2d40054297b35aa09639109c71fa78dad29ab773f0848ab85f8d919864c66,0b66f5f4f99e7badebb1e5f3cabbd1959c93fa7dc2fba067dede1169a5518be3,f61eb6a345f8238c11dfc6f1d56cdbc323a4376064207a4723ccaecafcc1b8ef,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,90.987,91.0,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,a8a78549c58fa3f3531cf28e78919807409f5bc933ff9642fd519bd94097142d,31842463221a1ac1961fc7cb477e8d41d2128fba686203011147dca1a88b256a,3829c5d77dce1f2b54d16d58262aa94aa7a4b31971c5e2b75494fe3c424c3f4e,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,95.55555555555556,95.6,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-08-06T23:59:11.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,frontiermath,FrontierMath T1–3,mathematics,b67c74d65e41a0617ff657060382a2ab6567db832842d7a978e5fae91889e360,13e7bf4931eadbf0eb78cc869e4825682ca4505ab096e335e9bebabf8d1656c0,1d7ed5f466056f5a6721a1fbc62165c689ab91d37b31c4cf255aff8f823bcdc4,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,59.64912280701754,59.6,exact,canonical-source,profile_87c06704555ddcf190611c15,reported,unknown,,unknown,,false,2026-06-11T23:35:57.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,arc-agi-2,ARC-AGI-2,abstraction,a05ff273772ad665615f3f3ecc6115a6fd716ec51391b25a8581520082caabf4,bb4071c7f5d7c5b3658854d5e63d4aa94083ebcb447ab6aae206a9a9a65c53d6,b51bdc08410f3edb3900c1005f74a21aa049ee81af67c033f943bc9ea06023e3,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,77.08,77,exact,canonical-source,profile_124badaa245c24f82cccc0ed,reported,unknown,,unknown,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,frontiermath-tier-4,FrontierMath T4,mathematics,be4fd80f5d20f319ec82c7f462bb8c8562c0d8c7e563da28f2fe9c2c97348426,6b9cd7b9eee50b02a7a343440113ce3c5c701a39eb0f701d5185d473bb3044bf,b2443e68eed4f08bf6e45e23dd0ebd1e0e724a61ee6a77c71bbc36ffa584d871,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,26.82926829268293,26.8,exact,canonical-source,profile_5baafa5d62b0d42201016d96,reported,unknown,,unknown,,false,2026-06-11T23:35:57.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,mmmu-pro,MMMU-Pro,multimodal,aa5b74a3f3c13307dd4ba3ce50efb73383a2f8f1955e8c99681550f4250d846f,8a0b4d8a48b849935fd9c59e4a829dc248888587cb6efd2b62f9208b14faa4ad,2b0bebc32c45446e79f049e96c754071a130f54318481baf283cb3fff8391b99,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,88.208,88.2,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,critpt,CritPt,science,00f43ac831113f91c6107180288ca9ddb0a6cf0337f7c719ce2954797cec9460,ef0696564ed519765a364150a449ef400611f0825930c26ad116b79ba690c788,e377e43859b258566b6306e15c3deffa6499eee502e08df18f4859efc44d459c,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,17.714286,17.7,exact,canonical-source,profile_8bf7a013a6ba02a14d73d95c,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,simplebench,SimpleBench,reasoning,7c6d36fc5dee8f9b3b033be17cd95b3ecacd783ac5050dc28985e812c3ca6821,13e48fb89b5acde4332a32d5a56e9b45b5d75281a3d2766f4e0c7b2a2948f27c,df2016c95911d075da9dcb44503a38e4051fe0cd9bfc78ecb819c4b5c28303cf,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,79.6,79.6,exact,canonical-source,profile_ff8f9f6ebdbd510dd6741832,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,chess-puzzles,Chess Puzzles,games,e7ba7fb66294642cd726753e9ffcef1b3ff7e8c7a8ef563f760368ca5e529019,ff3c986532d543608788fcba39c1b8421576cbc7670f69ed465dcc673fc40d2b,5a0cc8fa1189cb15473e5128726cd2235a11a28372026a0043df6363f456f2bd,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,51.0,51.0,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-10-13T13:34:13.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,arc-agi-1,ARC-AGI-1,abstraction,59fcb9caba86eb06169d5f4e374f27d6c2867ba5a60499d7108e48329a7ef6ab,300ae4d8df78d5ab7193c769903d5713a81eef3680a0cf122ec94a0abab646e7,cc2abe7a87614a9ab147026b1341186e1362cd461e7cafc629916f4dc8517201,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,98.0,98.0,exact,canonical-source,profile_fbb279d340e3a9059784a4f6,reported,unknown,,unknown,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,swe-bench-verified,SWE-bench Verified,coding-agent,4e427e30b1a9b19cb854a137f9e0acedb80fb4c6907c1f40a3fc3a91b15e1503,7ba58ea4d68df8e3e1f21cff5f1cff57c41d8ee2dc22ec84f84784a5a9af3f50,2a4f62aa36b6f99771eb16c33e2fd52bdf47ba7b3da8c214ff868379d2fbe719,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,78.8,78.8,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,swe-rebench-v2,SWE-rebench v2,coding-agent,f09c57a8a0af2f2cc0bbd086c5d7f44b8cf80d83d1d76613b8b6364cb86926a2,564af9ae9f84ba108ecbfca44c385f8016673f5283ddd1a26a20acf87b010202,264a1b30ff55d0e372595d202bedae349375f44074f5a61627eca3e3dd7d29e6,release_72d764965dd7baf34f983249,rolling-v2,2026-03-01/2026-05-15,resolved_rate,percent,true,51.0909090909091,51.09,other-supplemental,supplemental-prior-window,profile_2f6483fa19b9e55f4f65fa80,reported,unknown,,unknown,,false,2026-05-15T00:00:00Z,fresh,170,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,deepswe,DeepSWE,coding-agent,24531e87c0072f7ff251aad47f74d4c8a8d1d9817f4fa9ec9a809bf051d0107b,856701cddec8a0a373d524fdfea4bece9abb2064d37b7c6cef5f69d0b5645322,2091ed8cc90c49695eccd5e5a9623bb42511a74e8bdc6eda0bbd3d1d5fcb0f69,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,11.725663716814159,11.7,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,gso,GSO,coding-agent,ce1aebccb160f20d620875cf6ce7c7d2aaf50f2d749b143408dc53aa27513a2a,dccf6e263b7a7af27c0f35ed2c2713f9bba1de8232048c1ac2ce1e3d367946a9,b135f4b467d586011ba96e921074bd22ba5666ca1223d94e18542068d02118ae,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,22.55,22.6,exact,canonical-source,profile_73862c6e38001c572af5526a,reported,unknown,,unknown,,false,2026-03-09,stale,237,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,ioi,IOI,coding,54b1f67f6b9c651a55ee1ee80fa99e361d0dc4e686b793fe5d42518499f8ee30,6eefca776f67bc253e703c84dfa92dc169ab5410f69a0b4337d1e8fe259801fc,ff4191ac5fe3f4273a1ab1bf7bc5ed19a3b0912bfa63dc9c21c9f1604b16d17d,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,51.833,51.8,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-10-21T19:24:33Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,livecodebench,LiveCodeBench,coding,ac927f40690439bb8d8b5b7961b89f21a6bddfda67645568f545530254fe5760,64ea3bf4bcd060d82f94ab733b2fea4c3ee24df38397300f8152966062811fd2,6ccf48bb6200d6f017bd2d2ff0f6ce5508368543dc925f86971811b23a95f134,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,88.485,88.5,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,scicode,SciCode,coding,764eb8587a0612ec471ebe4250efaf5e8bd11df29a57e79da578a9a07a288ddd,fb7c0a683c8fe59f79cb4ac183565bc42a8a9e77fcb1f888f348d08ef8c4004c,e19ec468c2a5d484361cdde2c5ae2d160dfd77b7b5b9c540459be845a530fd7a,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,58.6805555555556,58.7,exact,canonical-source,profile_b7b3ebda2c95fe4a3c175a73,reported,unknown,,unknown,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,ale-bench,ALE-Bench,coding,eb8c37b674b6160674e3c38dcee6d23382313c0172324a97f0855ecbc2f8850b,0ceb3afeb1a724c9eaf1af61a09744d917c47879f52fa3559be12810ff1a9b45,bc779e3326d0bf2798439e12ee5b09564e180b02c37dd637e5d73d8d1578620f,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1160.6,1160.60,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,vibe-code-bench,Vibe Code Bench,coding-agent,4fb3fb10439b02bce962d8ec6042dac0cfc933296f93d149caf9c6c022ad7043,0beb942e32fc92aa254f27490702c1edfcfbd1e1c65ab54f932b3edc01b960bf,6033a7a00359ed4e825c4bcb86393e9401348b1cf80efccc3e31b93cb181d8c3,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,32.034,32.0,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,webdev-arena,Text Arena · Coding,coding,4ecdc3a94a5b81ff0015dd6389aca1f988bc0aadaeb3da38fd59a224be14c053,34cdb2a5610eb7e097837561c5229c3f0f35a3e2e6e7c58487ab680244f8f873,19a1dc6556160885f4ec8dc4ee56394cf69f60b22134da74591a47349587a662,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1446.6498483953492,1447,exact,canonical-source,profile_f3c287b85bdf2bddbf2899eb,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,mirrorcode,MirrorCode,coding-agent,418eba716f67e46c17fc0baf37667b96ddcd73e52fd6563ead81220ffa742042,476606e7d1c7d22e670da3783dd7028d922780e59ce52f86c800b3da64aba0a1,4060ee56d737012bf0bdb88584e8d4acdfa2d0d138dc49898229b803c43191ea,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,8.888888888888888,8.9,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-08-10T16:12:46.343Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,programbench,ProgramBench,coding-agent,97b1e9e5172321a118c63a752c6f47a5af5d57864f03896ac407c7a602a490a1,0e25fe3bf49e38ece61a92656a169b30bf970db3c80fe7681152edfc973b51fb,6c36c13e3d5f67c1ec660554f0f7faaae00eb46d2441ee4e5223b681b89bb60c,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,3a18a26e4632cc606b646dd510e64b4ece6ead7ce2345dcdc0488a4c1bd9022a,9bd5f17f8f943b05eb0dc3b21d2187edfd58e3bf2c908d81d09a7ee9f8336a77,59a46fae927a8ba671ab40cf6539bbf614163ed4b81f17aad271f78627b0b1bb,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,70.787,70.8,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,d2a817fbed4e5c86591ef775a2a4ad02a10dfd07c5e4aacd3a39b6a7ac172187,0fdefe70891cb29e05783072e513fcd0a9bcddf2ec57a16421bc79da8a468c5d,73c80a52b587e24957ad656ca4b84c5fde628b74f5dca87a57e00177f9d5ae4f,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,2.525,2.5,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,sre-bench,SRE Bench,terminal-agent,6ddafbc392a2ff2291d0b73ec9aac13471ad50a23566dd8dee6c1524fd63adc4,cb048da02183c6d8f0044b39235b07616f97ea4aa91c0f5b427daf0c8d4f586b,6682a12127eaca1f8518b02f5a574ca9b00531ccadf708490f448d4c9560155e,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,3d0e6b79a64e3df6f0fe4abf0db1496e8f51a473a61f190812f580df2219d7c4,46b4869dc80b84fbb8c0430435a8fc8767700d501f8887ed4aff5fb0b004697d,81168e94ccc34b0338a17620240d815652317bdde42f8682ad45a7b9e35df5da,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,21.443299,21.4,exact,canonical-source,profile_8bf7a013a6ba02a14d73d95c,reported,unknown,,unknown,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,mcp-atlas,MCP-Atlas · Scale,tool-use,943475666028d236ebbad1299561fd06113d1c578aebf60f40da5308228220f7,27c1879fe19f2fda4f4a0c3be591e0aba4d2499b7c895d8fb35012096f75dc8b,dc8cbb1b9e771d0dde12f7523330bf2cd18b1727080109019bec02d1199535e5,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,78.2,78.2,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2025-11-24T18:59:59Z,stale,341,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,skillsbench,SkillsBench v1.1,tool-use,b8796af64635661398b904d63b66723dc1024b861c50a18b65ac29e8056a40a5,cd788dad355a4a048591bb5e84b111298c20de68aefde7d8dae02cb0ae260e9c,b364c19f09c3235a2849dcf46f5f7497a801c57e4a3db1ecf5efd428de0c8418,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,60.8,60.8,exact,canonical-source,profile_522565efd30b538ee4ac5d42,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,apex-agents,APEX-Agents,tool-use,707804e227dbbac6fece776b3c3e76d75d4ba20b612b89e2590172b82fa980d7,e44ff48b2657a35cb5a46f08140576912d683a64b6791f3be75ab848a622f4b2,fe4e4e4739bce6a15ce7acd0f662b9c155d0480a074e85d6b6ecf6494498998f,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,35.1,35.1,exact,canonical-source,profile_3ad665509e47df167de108cf,reported,unknown,,unknown,,false,2026-10-19T22:24:14Z,fresh,12,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,arena-agent,LMArena Agent Arena,multi-turn-agent,4ee1f9d079117d3e79e59f7059c688f3ef808123b82b1c6c3cd0efe274e4d69c,5652716c16ca782b6b6cefe47a26774290cd3bf2f3b793ac18c9297645d49fbe,6f94dfc35f929e177e9f9b1c5b6e04eff77cbbd32e74cced6e05c47a5e5a97ad,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.044514063835776815,'-0.0445,exact,source-specific,profile_71d78988738e81ae5ae5bd43,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,vending-bench-2,Vending-Bench 2,multi-turn-agent,17a77adbf1b24556ecd1d78eafd3d6b095e32cde4fae49ca25c8c05adc1f6f9e,54cb3c06127e67b09523262c124d6cd130271ed4dd06553cb1668697e10fd161,8259756d9448652adb9c8e2ee1fe06f67271648e2f5f364b0f1709d484d02ab4,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,911.2139999999997,911.21,exact,canonical-source,profile_e91ccba2908d54447036bd3b,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,40b5c228d92595657767a8766f46be6e53eff48a42f3638013be7f5eda772096,dd714c184a8d3a1550ca5c38aee8e5e7aad9c898331faab26150cbae5d836697,fb609b4a1cecd9a14192ad6c2e961133f3c72b8c9825dfd3c68a77934bbbd90a,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,798.88,799,exact,canonical-source,profile_8bf7a013a6ba02a14d73d95c,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,9debe32a3face03ccb1e208278f66e81dd16864770828525ba0c5c14ed9f11b7,5f5cbc0486144d76d1864d11946e34af8a3ba69ddf6677d51b7ee3cb5e25921b,0c9950239c54bc8946cf37a8431907d28f566bdae18d7b250476128e7b838fdc,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,1.429,1.4,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-10-26T22:05:28Z,fresh,5,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,070142ff00f7d669c7037fcd12d788543c794206ecfbdee96c63ea7b068fd90c,5705135e7e214fb580ffab2117b7d06653bb4b7384095c135e414adc6bcc559d,cc2b5f2e843b4d0088ddb39baa3601946a893dd4972691552d040c8ddbfcbdb9,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,26.03,26.0,exact,canonical-source,profile_dd3be8508497fcbd983b0185,high,high,true,fixed_effort,,false,2026-05-05,fresh,180,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
google/gemini-3-1-pro,Gemini 3.1 Pro,Google,closed,37,weirdml,WeirdML,coding,7fb7aac72716d54c205f47de179ff98395bb89b038a9d9999b4679dcba30946c,591741da0aa50b36cff5d925a4458c59efed5e22f424664b01236d81d71a9e57,2d55e99abd048ba98d8233abd81925bb13dcd26af11a5c45a211addf5d7e63d1,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,72.07,72.1,exact,canonical-source,profile_32dba87df2012b02f4613fb6,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,epoch-eci,Epoch Capabilities Index,capability-index,d3161afd35b4f8055f497e73f407d6c8ec5fb2a4d3565a0b80d369dee64dff84,21a37dc68ff2b083d14bcfedb9a27d112e6975b285dfc15efb9c4be32d4e02f3,897ba1cd014dcd819dd68489041b9f035f5cce425258e9ef0e38e65ece4fd970,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,155.2,155.20,exact,canonical-source,profile_07d1816d63900bef656c0e24,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,arena-text,Arena Text,preference,5b05861534997157de975f487d06794b181a3a4308eb12999c0d18a7dde62435,1bd3f35fbf6c34d8c0d7bad5abf52816cf8046e38bd40dfd12494c7beec4e2b5,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1492.188040510066,1492,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,livebench,LiveBench,general,0aa93282e1af17cddcff8520264c7c79de6271bc70ef28c1a33a4ade657c8a50,022137b4eb7ac10ee27397a97b1c44af4f8a81a5be6bddf6d84a62fc861fbba2,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,77.9547738095238,77.95,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,197ef60938beb17ac0bd2729750ca599f22700566555c68c6423c7210d8006e9,554e935bd6f3b4fa060b4263f50fecc07f64f16dee7eacda35ac3772edacc6cd,b9554e2032c7be8036d239d362912f1ce3541c0223e284819304e8efaae5e25a,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,45.458758,45.5,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,simpleqa-verified,SimpleQA Verified,factuality,f0851812d1d42546731e31e6a0d97940d196dec2592905584db900de801cca19,0c2f1e0739e98bae3a2a1ddcbd9cccf6975f56b01a6124dcbe45b837627173b6,cf5cdcc4c3a70b7351aabe6cbba24f9eca000dac537b1ffcdeb66a58b7031afe,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,60.30150753768844,60.3,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T19:36:49.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,gpqa-diamond,GPQA Diamond,science,bb25b30e4db8b40325a2ff2579708175890deca8c7faeb87875d0e26c9494f1a,fa240e3bfe3d08d0494982d85de1d7f742ea7b7a4518d5ea86b8cda5c37be6a1,87f968a546c9523fe5682c528ae62c619eecf75dda377913fd1ef5d1055eeba0,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,90.40404,90.4,other-supplemental,supplemental-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,mmlu-pro,MMLU-Pro,knowledge,1b4186d450816ff4d658ef6d59ea3f38885db8d68eae4d3f77d32ad2fd9f9c62,b38e45c85d3e4add28d8ef84b76d48356aeacc355a73767874355308b1036368,29fcfe5a022d697f016363a005652aba0cfb4c9cedef845707258e2657b850de,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,88.275,88.3,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,mmmu-pro,MMMU-Pro,multimodal,48b266ad07f757bf2cff3eeba59baccfc5b08bffacf10e55a661e48c47785ba8,cff51443d433ddc78bb63164c02d0127d81fba05a4ac0e2d0e723b10dd884522,108f7c7932ab215ebec5bd11827b1bda1a54703db7eb9b1ca54d66dba28f97fd,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,86.127,86.1,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,critpt,CritPt,science,a418a974072c93df0aa9928f6ae402f4b4dc494c21a09149de79706ad426ad98,0caaa8b181c8f3685dd47ef28e1356db69aed700747aeda739e94cb997b14d1e,f1be2f82d6b34791f82dd4f67f3085e36bd1526bd543a0931c438f9fa63a3c5f,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,17.714285714285698,17.7,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,simplebench,SimpleBench,reasoning,104c2f26af251abd3677f648b7c9cec6bc434ff9b8a6a8cd8fd792e1a9d91f58,82c0b3e69c46fc148ec9291ee316740a53f7beeba2277aa92407958d3049b2a8,84da92f7c5c07b1b3e560be218e49d78ebd66a5ca25cdcb81951f69bbdb448f2,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,74.5,74.5,exact,canonical-source,profile_01969cbf5dae3b1269d94614,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,chess-puzzles,Chess Puzzles,games,3e6b4690647bf9f5fff9dd6097504f0b8a216dbce3d3ebd332ca2b9e58cb7dbb,407a02f6115c0c46d58ec28cccfe80919cc1bbcf4313b3f58c971395014c332c,34bd3c2a792cb1d370ccff4e8e6e03902c73685013d0a23a18d99e26214606b8,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,27.0,27.0,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-13T13:05:03.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,swe-bench-verified,SWE-bench Verified,coding-agent,961b85f92c9f7e2f85b7ab757b8d581398c32fbbc7d78fdd2fa99cd6249b6395,2b13dee71ed10880fba26a53493de1c7d0419be5a72c128a05f7870d4c99429d,a05b09281606b4ab4972d51b72d9530d02afb4017a66cfa2b417cb933abe8878,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,86.6,86.6,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,deepswe,DeepSWE,coding-agent,e3d8e74aad1585cf832823e02c0f2cbd6e414cda49c90a5544288226891aef24,6f36446b918ecd4442e0c62dccc20ec6009e5cb2005585e6294c340871628c66,28282cf783bafcecd7f7d9d45b49a4ca5f96b889be7c9be1937e83965b47aca2,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,54.86725663716814,54.9,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,ioi,IOI,coding,3eaebadfec04a6437b61d64e3d54aeb4e672973afc1cefef5e23bdd41a6acd95,8449ecd9dce4e2c1b0ec5bd2719f2ad05d183f739d24f9aa261963475f0365c7,bd6f744bd3c5e33dab303062e574b07c9a19d9591f9ec619ca3a7bb0c7cd049e,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,21.778,21.8,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,scicode,SciCode,coding,ac3e14a72e53eeb2470e56db37f5c56f71bb1a6e08bcf3ce3df53a857b5e725e,3d86fd25f7b894489ceab3edff05f505277704e05323259475e0ce4bb0e942ce,969fc0db003b7c4854887ab612f2d7c2560c84785d1535ec2b08f7c8bb3da7fd,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,57.4074074074074,57.4,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,vibe-code-bench,Vibe Code Bench,coding-agent,c16ac1945db665463ebac5e3d3766b28b7d9a265f61032cad8dccde44df0f31c,d6b37fa31ce2d740db201b8fd0cecc3da9e1f0861d9f753c1a725dbc0be39359,067caee6503667936eaaef788d6dc1e1bd09f23e7a74d02c9ec79dd17d1d84f5,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,79.1,79.1,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,webdev-arena,Text Arena · Coding,coding,985e598ae83150f6779bec51e2bca1f54790a3f3411bd55a7f7ff9d044fadc53,8b7f0331cb926e593003bae6df4606c8992c6b1a53ef831f6d65ed3e879585ab,bfddb5d36dacac6f5c175a6a5a714a48922d75d3caf6d966d8c825edda7352e2,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1533.6689783572438,1534,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,programbench,ProgramBench,coding-agent,90bdde53e0da7edbee24046dbc3fb007c86c07b750bef0aaf7fd4f10c6d83e02,57a69771b6d12badbcf334ba56d6374b638299b268f8e9f446c9a9a63af83f0c,567326fe8a298690587f810783b49500f521a8f33809ac31d1198a98dfaa6690,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.5,0.5,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,65cfba9a9f0e55e957b3fc8c1fd145304240ec461d93455b926e8feab197ba50,d09a08ea556690f157e07eb6f2614881c1bce10bed7d21b1d7ea76daef217139,710a0843ce8d4d01d7e65f58c3ebdcab4c378674ce49fcf99cfba53ad4bf89db,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,69.663,69.7,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,d41bc6ff863adbb5425c82089fb58b03bf432e81bd8d239219f120921eaa42c0,5bbed6897abb94628222732fe99dbac31f853e9f0d3a26373367029e3e1b706d,92883ef74af8abf93fdc22edd8361bc0ed1fdc5d5e003f1d09fccdf02ceee695,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,6.061,6.1,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,sre-bench,SRE Bench,terminal-agent,65aca63e7eb0a00a52a0ff19c2d18bc294bf722f569ed2cd230bb9e2bbb32728,2977bd730a2c1f1a6a53118546eea0fc8c38f8734176a7af07ff68f7f3c5723d,d70054ca3eefd85179d02ea87d444bef70efc3afc35359f41b937a5fcf69938e,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,863f3f34164994d823b7ae98e1c59b47ff2b2c1dc9eade7cc0f02566399af225,48b525abb074718ab62c6b1902ed78800280815b08d0ee5028886f67ba4add02,60a5c6f1fbabc90023fd5a9c18df1c000d431d1e29a2930a21405eeda84e113f,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,34.845361,34.8,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,apex-agents,APEX-Agents,tool-use,93810fb5e60b31f8b64419db95a5854ecfac7670495e6171104a039df73a7b38,20d195a45757e4e4eb4fb4c6a4816a76943af125432947f16dc04a659ab9e278,b07d59e3860875786cdcce441391a5f01684b899578f4e5e3abe5ce59f889e02,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,45.4,45.4,exact,canonical-source,profile_6949cf33fc5bd1ec3a3284a7,reported,unknown,,unknown,,false,2026-10-19T22:24:14Z,fresh,12,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,arena-agent,LMArena Agent Arena,multi-turn-agent,15bee127ea2fffc10a2502cf9ba23caf541a0a7f583a5ea21e96c0bdef73323d,1b792183515ab5a456f3d109c9d79984fef8bf965f91760c81521695c4d7533a,66f20e294edf67f19528ae22d5ee44e90cdf0a74c8e17cb02bbebe5939271f73,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.07538020659982388,'-0.0754,exact,source-specific,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,928cb834c592733f51a4a390ac2fd58ec836aff6b33dda038d3e498f703bc66e,23b91385e6a6c333cac4582fb386731afbed16700bd125683d74f0a79160cd4e,370129e05a5ccfbe9a4d83e2cc78db918a46628871d8a9f6ef50b427589db98b,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1480.04,1480,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
meta/muse-spark-1-2,Muse Spark 1.2,Meta,closed,38,weirdml,WeirdML,coding,cb5235d4127530b7322aa96a7823873b9c3e0c8382b0a7cc130d513d900a6897,e95b9b0cbca4e68a6fcc162391bdf5ecd9f197f57c31727f667332ebf211d9c3,40925098be8d19a7ea155e1d80d593ef09345a8b8ceec661dc3c16dd00655b43,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,60.3,60.3,exact,canonical-source,profile_502229b22892e7bba82f7d86,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,epoch-eci,Epoch Capabilities Index,capability-index,5c79be64953c23d3a5309f3c2dcd3c612af530ce5fa26890c0b73f8f754af1a5,bbaeccf3ebd693bf4614f8376b4e475d9a7d91f10a6a5eb04466c23dfdc20c18,5badc9ecd93bd2ac2560d010034e9de572b9eec872420d4878ceb2be59d7b35f,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,155.05,155.05,exact,canonical-source,profile_9178b0a2d0e284b4c757f3e1,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,b9fbd3fe77b021891d4eec2b4f629d8a8f4b6abdfa6126d31064b80b56331051,2d61a00c4e1041e76ee7ade9627ed829213b2f5239339a41d6763b80ab498add,2d9d2214c3d6f813be698696200aa143e893347e9c02ab7fc79b26278a16fe62,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,24.681587,24.7,exact,canonical-source,profile_ae2b0889cadd601ae1049f93,xhigh,xhigh,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,1aaafae53b446574530c0ad9989072441a049f079754414f0fc00ab181e1d898,ea6fdd5e934986c960ca595c4f7880e798321e9681841b2b1044b47b850e9d79,8e78c908cdfe6241a6af1211b300c2684969f5ab8db4e3024a351fe70c7e69d8,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,43.883225,43.9,exact,canonical-source,profile_ae2b0889cadd601ae1049f93,xhigh,xhigh,true,fixed_effort,,false,2026-10-15T22:20:56Z,fresh,16,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,gpqa-diamond,GPQA Diamond,science,7d6d620af504f71fe589a413fa7c8df838b4be34e97315f0ba89df02a2ffe921,e954f4c3438a009b34918e9413ff24f6fc22e9e2b8fceb4cdde4fdf814039873,02945c406adec6e1444f7b02c388dbbe69ed29a02bdd52ddcdf480b63a43f5c3,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,88.989899,89.0,other-supplemental,supplemental-source,profile_ae2b0889cadd601ae1049f93,xhigh,xhigh,true,fixed_effort,,false,2026-10-15T22:20:56Z,fresh,16,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,arc-agi-2,ARC-AGI-2,abstraction,3376ba48e508af44878ec9a45ddf8a961385e0ae7c1f92b948512f64b1cce29f,2430566af4b3bc27706339c8071b027b4a2c0823eac198f3e140506cb2ac8a29,2dc7d3cd93c2288c1e025b143d853046083ac506f18d333af907400bb534d896,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,57.49999999999999,57,exact,canonical-source,profile_ae2b0889cadd601ae1049f93,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T22:23:16Z,fresh,10,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,mmmu-pro,MMMU-Pro,multimodal,75d5d412c5864c1876dd07b476a319dbadb0d3c73b11d80cd68d3a13d9062725,4ff82e59f0a4efb4684261f27d2e46401f288f6bf17e3aaa2a57a6b4f9986f4d,904bf5b0dd6c8ba039f5ed7f92a770f585817f587d0143ba195f70006316a4b1,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,80.924855,80.9,other-supplemental,supplemental-source,profile_ae2b0889cadd601ae1049f93,xhigh,xhigh,true,fixed_effort,,false,2026-10-15T22:20:56Z,fresh,16,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,critpt,CritPt,science,549706da91d516ac8e0cc119451f2c3a8b6761115a3d7add91f00d5633e39801,fd7147f6a7783384f27b7bbd89af2187ca2e687fb9ab73938d4f0d6e45d9dca7,a470794991406c565e1aee3510a0242e5dcdcb7aa7c2c3630fd6fa353504908f,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,14.857143,14.9,exact,canonical-source,profile_ae2b0889cadd601ae1049f93,xhigh,xhigh,true,fixed_effort,,false,2026-10-15T22:20:56Z,fresh,16,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,c12da0ad2b3939c34b50adf4e6662a100407bbcc19613c073c8a6ba2b6808572,7983b9c79bd562f7067bfc9a160e9cc288fa045812a68204af646a0a1fca55a3,7260746789a5f91aa487f5a7b2367ba81f6bf0fba1a5554ada113bcf9f9e5a27,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,95.02,95.0,exact,canonical-source,profile_ae2b0889cadd601ae1049f93,xhigh,xhigh,true,fixed_effort,,false,2026-10-16T19:17:34Z,fresh,15,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,scicode,SciCode,coding,ec258ed81fa6995852519b76d26f6aa15e714daf6507d46ad6ac97c44e4fc3e2,81bda2856fcc90761ecd6c3e05a405e0f5427ccd168dd8f38e52a9b65ea83d81,c8da256b9b53f1d1a24afcb996406071609da22730ac8cb061c9f886c06a2a73,release_de40c66bd40a7b5483b40f9a,SciCode v1.0.1 (Artificial Analysis),scientific-research-coding-subproblems,subproblem_accuracy,percent,true,58.449074,58.4,other-supplemental,supplemental-source,profile_ae2b0889cadd601ae1049f93,xhigh,xhigh,true,fixed_effort,,false,2026-10-15T22:20:56Z,fresh,16,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,vibe-code-bench,Vibe Code Bench,coding-agent,c223fd7d2f09dbebd6d3fbd50dbf483c4ea73e244be385eac8a8fd763215842e,9caf2221f1b021f9c31a199a4c71f17fe3a9763f8ec933e4a692737d15b3d497,b485a060cfee9de412985e7facaafce4f190c2e4b4a248ba16acc26ef5472f13,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,76.187,76.2,exact,canonical-source,profile_ae2b0889cadd601ae1049f93,xhigh,xhigh,true,fixed_effort,,false,2026-10-16T22:25:04Z,fresh,15,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,webdev-arena,Text Arena · Coding,coding,76c2002e70f8af30527510138d38d9ffc8a2cd345abc7f302cbaab90e726d8e8,25d1a1b91535e325bd3f41af1c2e64c0c8cceec493a3415e4912f383076d6344,10722bd31ad4a3fff0e41aec9037abe13451c8cfcb9dc04feddd0439d29e0259,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1526.2373165722156,1526,exact,canonical-source,profile_08650aebe41ef808d3d674e8,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,9af7042cf7973d5692bf8ecb6b93957925aa9a442c3e0a949a18f9f7b6a6f71b,820522ae6eef6bd7e97954e5a0a23c4b432da4af781fe08b8c27af2150d86943,28b9626d77409af07f48a97bda59053fc96928ac1133a83464d1e96863fb7ca8,release_c53c830678318dc1763793c8,Terminal-Bench 2.1 (Artificial Analysis),terminus-2-89-tasks,pass_at_1,percent,true,76.404494,76.4,other-supplemental,supplemental-source,profile_ae2b0889cadd601ae1049f93,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,90fc7e536f90ef1d799b3bfaaeef13d18a0007d7c9bc2672383134266c27073e,ca335c6282ef2598fd276db5fc9ed2b11895e66609559006df8396e24cb2892e,2b464bd20d79c0d0a2f19f2cef2f0b3b2ea518802fbff3c9e8bb485a6340ed56,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,6.566,6.6,exact,canonical-source,profile_ae2b0889cadd601ae1049f93,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
meta/muse-spark-open,Muse Spark Open,Meta,unknown,39,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,c8081ddd4ffdde6a6f95ce1249d506ef6495a1075616c20dbc4fda9917416b41,2fd20ce0ea2054a5e56716fc42398ec7921b1ac2a8f9cc01b52727eaee4df795,ea180ff663f5ccf2268ccf0375d2e297501595050a9c5bf3a2e4054aa9aa34d6,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1401.12,1401,exact,canonical-source,profile_ae2b0889cadd601ae1049f93,xhigh,xhigh,true,fixed_effort,,false,2026-10-15T22:20:56Z,fresh,16,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,epoch-eci,Epoch Capabilities Index,capability-index,c570ca6a809ce2b39f0df9838fca6aaf6104aef88a7384e003b5acc396d9af5a,ef0b8a2dcc41cb1558c5db33eb626068d086f4561d89300fb350f5bb98bcb755,8b6b6e08dcbcb2d0e73673fb8d80971b11af2cd9300ecf366870494c77a14a7c,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,155.03,155.03,exact,canonical-source,profile_d3b2abb9ad25b246a9b715e1,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,arena-text,Arena Text,preference,9d6cf1d0710548e16a13eac07212d2ff95fd040a5352501d310a05b417b8af35,cb3888e059bdaa4ae60bb47ac333fedb2f3f23b10f18e7a00c834ccd300308a0,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1477.6273768777826,1478,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,livebench,LiveBench,general,6b7c5944affe78ff9d68051b94db0eff2dfbf7cd989949109f48da738dfbfd79,708ed4ae8d352d0055ca44aa7719a2c8139ee77bb1f55b8248ca0b3c3c869cfa,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,74.6377619047619,74.64,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,797fc2e0417b47624835cb25f38009e532d955528b95007dd153ca50a8b0007a,a8b4ed847af0e912a3f44e915ab3525871c03bfee7594d7dd8b12ce74f152abc,f1c14751cb135ea7742188a867c1ba4dd19b3da37521600b924d9d938272dc0a,release_5f3023448c2ae4632fe1a43b,reported-runs,public-multimodal-mixed-run-configs,accuracy,percent,true,42.5,42.5,other-supplemental,supplemental-source,profile_04a576d1beb0fd7c718f6932,reported,unknown,,unknown,,false,2026-05-19,fresh,166,2,B,2,B,1,A,evals-report-official,0,evals-report-official,evals-report-hle,evals.report · Humanity's Last Exam,https://evals.report/benchmarks/humanitys-last-exam?tab=scores
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,c0fe6a9d71f22fca1b8e9f9038653d0cba34d45f1afd191e046d2fcd66c1d612,963f4bfa82beb3aae899062359d6ccf45ff8cf09234f8a4e42547b2c1741eab0,ab03824db671c6d94897a344c9a9308acfa6123fc7a3526b1403f986d011dc68,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,41.334569,41.3,exact,canonical-source,profile_667b754faded84c3180149c6,medium,medium,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,simpleqa-verified,SimpleQA Verified,factuality,eab486438c0bd1641cff4595132c722797647446283b6aafd36c8d801d5a9651,39a4d982b5bb9b0d31f6f1e541525de71866abbd2a7986b43fde60f4cdc9fb6a,e862642b958218745c8c4c30f3ef5e81069ee89a2112ae7eeb5a3627c1e7c8a1,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,66.2,66.2,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-08-27T19:30:35.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,gpqa-diamond,GPQA Diamond,science,6fc4d9737a663634ae2d0a0612e91e720c076499a210b5dfcc0cc273127749a0,c9a84746b84ee524e51afceb13b0449cf63a06fcf510c65dc26a9f5b76c78e6b,304c1cee578e771901d2db02a68e8709b0786b9de8311b3726c1a0d9377bc126,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,92.676,92.68,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,mmlu-pro,MMLU-Pro,knowledge,4c6dd49dcee6ec97591fd4261791ad4d7eb04711b528f4a4a379505941d2172f,795a9fa75592607418a47e300a4dd545ed9e72dbf95aa0637ebb3a0c6b2fd0b3,c99c68e86d5603c9191cbecf32c36729901f529ecc04ed77c0f9c22dcb08473a,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,89.515,89.5,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,c4401c6bfe9849bac82782b692e77dfc59427ba60dcaa972e71d3d135b2e31e6,d7aefe1534c1c17f4b4a17fb117c903731e37c96dfa26a127bd1f9e6cbfc6e88,f5fdf83eb3d4f1fc9c2086fc3f859867e4070da47236d1a1788e969ac7c2d436,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,95.55555555555556,95.6,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-05-25T13:10:18.000Z,fresh,159,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,frontiermath,FrontierMath T1–3,mathematics,2b2658e74e793576ac7e809c9524683e466eded96b25a7c0c7ba9cc0ee5f4919,1ef1a6ea05794420def5d7e2da39c5953159e8fc20d802ef837817e7ce8efff4,daa218ae0b434ce51d940177b4aeb7f1dde4b11fe1e481fe1df08e75d4d6e5b7,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,62.807017543859644,62.8,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-06-10T17:32:38.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,arc-agi-2,ARC-AGI-2,abstraction,0dd4461d31eedfc95e677b60cbba1111f07e770498a04b849e458f30c9b178c7,986e07d8e3f525c63994847c40858db2472b4a43024065b5676e0774a70e540a,dbe38c082ebd2c37cdc880c1706306279374fbeff0205fdba64e0d763061a9c1,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,72.08,72,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,frontiermath-tier-4,FrontierMath T4,mathematics,a7c1cadcee572b49b121cb141e325509adcdc9879cb3b3ec3a3e3093f9450564,b694cfd043a6c8b4585b08fdabee36f46867346dc0f19c361b6bed4d7772ca0f,61b9e7719f4d628014c835305095e2f34f2afa822735c4212159bb6315897cbc,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,26.82926829268293,26.8,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-06-10T17:32:38.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,mmmu-pro,MMMU-Pro,multimodal,7a9b044a48e37df0fbf1d320aa122dd8289ff39e10d5b081ce61bf1d9d1beafc,409a07a754f129dd45772147fb6e5323c835a3f8965c1318ce3c35246e79c188,d75a988a9863e64765f093d2198ebd74f41c3337b75c4e9b2d5587232c690d84,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,88.266,88.3,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,critpt,CritPt,science,7697da0109ce2fe06607ece192ceede7e6a35d7c513eeb2122a2e1548b346fc3,988c5973533a99a0123ef54bc81be6dc340d2c38c61e93c60ef01e6b48a7041f,766c1b2908381ba0fb0250e4b074ff1189922b8db55cc8dc5471587266323044,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,13.1428571428571,13.1,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,simplebench,SimpleBench,reasoning,6e8532cb8111c25943ffd62587cfd9e01b73ee5539c6b91ecc57736c13fc3bfb,cf3325512eb74ed46463d94ae554035f962e467b43c92fd8ef458b536477a32c,34c405ecef08f9fa68202ed884d799037870c7441b23f79acba78c81c9c5ff3b,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,76.7,76.7,exact,canonical-source,profile_712ce459e409b423c301b510,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,chess-puzzles,Chess Puzzles,games,e91ea39c2b330ef314e0d4ff3ff9633b88a7fdfef0e5ad2f84eb878f34a96323,9fdd0d389f9c264c2a613a2b33fdee7a5923c65cda035c49e737a0b5ff0ce47a,d3d9f1de2322e388d7485ed2cca2fb490bda7323295cef7b5f1de2ccd57821b7,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,45.0,45.0,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-10-13T13:33:53.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,arc-agi-1,ARC-AGI-1,abstraction,e53c40e679296217d4b462976a8ab03d700e14f958d12db534d4a9fae666c5ba,5b807803ec428066c222040ae61b3bc1576e4e0d15f2423d47eab044c64c6b6b,3194eb015bdbed7128bfaecdd95ce7a9afe9ccb8aada513b5e6774c14e91a543,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,92.5,92.5,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,swe-bench-verified,SWE-bench Verified,coding-agent,70a8c70aaeebe42bd19d5f7f3f911f0346a4e195b47e61c822f624235b191609,e787a6288f44471e3257433918aadca25cb7d0e4d60680434527a24f2b7b47cf,0cf1f0f1968fb021997b12c434a3e3f55eaba1b4024d6a2956d2143c7dfcf378,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,78.8,78.8,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,swe-rebench-v2,SWE-rebench v2,coding-agent,c2d9fbd00900a879634dead4ec872148747718dc01090f62b1e14637ad6879ca,aa675fde7b44f1357ce0d43de8b18ac3ed435f64c51df88a50a561f55ca43927,87cae4fc720c39d5f79492dbc012e5444b315f8a6cb11bd00d5822d440b577ef,release_72d764965dd7baf34f983249,rolling-v2,2026-03-01/2026-05-15,resolved_rate,percent,true,49.45454545454545,49.45,other-supplemental,supplemental-prior-window,profile_336dda2b1636ac9292005a45,reported,unknown,,unknown,,false,2026-08-30T11:02:42Z,fresh,62,0,A+,0,A+,7,C,potential-contamination,6,potential-contamination,swe-rebench,SWE-rebench,https://swe-rebench.com/
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,deepswe,DeepSWE,coding-agent,1b64508ae33f3005a1dfcc82f19a4a06edcd10cd5aad07c7a75af2b592b07ee1,35ff8572ac1eeb0e063ed04b1ee19ed55c00e59a9a51fca597972eee7b8986c1,e47ead2f51c572dbfc57ebe743690c269e3b2a60004b56d6c13a0123d91e6db9,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,36.06194690265487,36.1,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,livecodebench,LiveCodeBench,coding,5ab762cefd0315edc036879a3048bc1b262ca99504f1a24894f97967b730a46c,fa06f85e345deeb9f84fd5e5a4f4480ec908e1753445eb8dc6146746499246b5,75f4a201a33e83fcc7374d51a81fcf686107533e207e87f40a95800f0c1d555f,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,87.604,87.6,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,scicode,SciCode,coding,6f5ead39d832886a6da6c4fe9a5f30a3fb00a3b3c91600eb6546aabcaa04e8d4,63b08be79fa6b1215eb97869436accdcb902962658ca16a7de330f9b25625bc0,d8653dae3ef9b6db54408e89e286a24bf9e163cab4e5dae462577f628dcfdd5a,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,53.9351851851852,53.9,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,ale-bench,ALE-Bench,coding,64258336583c0e15d6f0766488b5a1cc54defde59e2ffdbf5d1ddf2250276734,7d6a3e3dc94a256f5d3ad7546c2970637f7a651ef7d1a0ad839b5cb097571a94,ff7c4e8d844244d6df0ad98ead4952ab0d5bf0cd9dd1e645c58dab28adfee8d8,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,911.02,911.02,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,vibe-code-bench,Vibe Code Bench,coding-agent,1a55688850e5d0779682855b841fcba6b71ff0c599e84bd06c3d38d9904a8676,43c363da0b1c4cb89dc96225f534836b4ccdf712496f8579c17ebf01f94561b5,29b627d5308e800ec60ee7cbb05e2e32ebaa9b69cc59a052fd62298d5d7878af,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,48.683,48.7,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,webdev-arena,Text Arena · Coding,coding,5395e723b3cd52f198ab738650eb2d2b17a6c4ec78ac1a1b8dddd56fb069d3e2,c7837440e1280dbe086f0ae2ee2cfcd0c6a412135fe4f93a8c7783fc08ad00e2,13b8f5367dbd56386a64d986c5b951bd9ffdbb656ecc58dcd600a42d18b3037f,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1499.2741854109668,1499,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,programbench,ProgramBench,coding-agent,781dd298a107acfab64645dcbfa8f45028903baa85a5c53395e7e4c8232c2944,fb4c7952c71b7ff55a4d43990f92b63ad97eb0b20fd0d2a4d67c5305a4659852,33d92bd143476db9eb9854c145519019cd84cd865ab701c9b8fdc4dc81ce85f0,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,bc40b059527f4a237415a587f6973d241ba71fbfd858fd5a5e66ad9067c69a80,5cb948c77f0de72b4e98a0eafdd3a3d50b28163c24abf51327967b7a6a67d69f,a014b01503734ff9164368c4f97d154c3388ce94c5bfb6f29d74eb3d75f6cbd8,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,74.157,74.2,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,e0bdd95341a4e6da7c95e3f98735df935184f39ff9f7d3706ea3772ec4e58b63,6622d6bc2d8c7f3884cc52401e07967d6a61645b0eaea1bbd20df02b7f71ae29,9e273746af5615484192acb98c52aba78998a73c5173cfa6270acf34e0069bcc,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,6.061,6.1,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,mcp-atlas,MCP-Atlas · Scale,tool-use,c26165b5a6c6e19fd9367e694bdfdc3ce069171cfd02cd3b26144dae886b1c4e,6c55c974142cb7b36d8703ace2c48a502b23728939f5e76fdbe096c3b97de471,3e221025507df5a388c40e3c45071624ad99dfb94edd96e448d2c144843437d5,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,83.6,83.6,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-05-19T18:56:30Z,fresh,165,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,skillsbench,SkillsBench v1.1,tool-use,8bceecd92221db77944ffda42e2ab85e21337caf3d2e23a688cb5a09e26a2599,8d8c29bcb5bf7f7aa3dc0d1cf8a5006f636be15f031cb5c09f584289e32a5962,c2ed2025e534826e5e627446f4493a128269710491b35eb2122abbaf19d56304,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,48.2,48.2,exact,canonical-source,profile_c2aed1feafbd96f260e04542,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,apex-agents,APEX-Agents,tool-use,c95d355c6880d9e12ceb443a8f77347500c8d0244f94965eddc0131c5f325a49,900dcf5a435675c7ac49286898781d650c33db0872cbd5d421937bef595cc20a,4476fc4b13aef3a80e395fd04a9a309eaf20f4322b18ff7acbe86126714a050a,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,27.500000000000004,27.5,exact,canonical-source,profile_97f8426db315c9e742f4bc9d,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,vending-bench-2,Vending-Bench 2,multi-turn-agent,c01beb7291e936b8dbae7a7f405f2015031fbf541a64bbd938a2cc26d351c7a6,201657be2e500acaf220fec775fde1af87601ac7b096888511ed94a669aeaa1f,aa8132967196758e9cafa32477dd5915441e1de5b9abbf10bab0fc900987ef2e,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,5396.418333333334,5396.42,exact,canonical-source,profile_81b485f877e695032b0db2d4,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,d9565715e1026bf2de367808a6559e0f463bb2055c550bf01e7be98425849c43,4bbde2a867f5b49ebb7ae62237e9ab094286ac85bee6838bf8b33db109e7f9be,3782b4ff4913ffae588e1473c4fb03ab58a239599887b4d8c34c43b28aaa1bb8,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1211.55,1212,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,539841032994970020f506c87353431112fc0829279c43b905f257a037c20143,8840c270f50ad87e8bc7c30f73bc2ccded842961f8acc0b1fa9f65a4ba54bf64,60de94512bc8d3ae20116fe763c1114e524f568ae626d53682eb1ba09013895a,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,5.714,5.7,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-10-26T22:05:28Z,fresh,5,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
google/gemini-3-5-flash,Gemini 3.5 Flash,Google,closed,40,weirdml,WeirdML,coding,ee714246d67430b6401dd16aeedbac5946d9368cda8b7a8a3e0154f8fec31d98,c779f9877210d4b36460518eb7b226770e9d907947d8b564af74578403b2c340,6314b3888394b0ce27cf106727cd174d83c0e0776ca36ea799a0535697dd4c89,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,62.64000000000001,62.6,exact,canonical-source,profile_eccca673c5e71c05cc78a977,high,high,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,epoch-eci,Epoch Capabilities Index,capability-index,9ffb968bc287919a163ce34e9924aa84950ed12a1bafa49e527ce5ed6e2a74b0,5dc8008b0b4574f969540f6613add8841b8c7494df00c53908c73368f591e68e,be3eca2869814a07109082268b5d094d1a1f909f86c6c90949ba8eea6f958b18,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,154.94,154.94,exact,canonical-source,profile_21b864923f5257b0d3f99f91,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,livebench,LiveBench,general,16040d54bdc0aa99749366be940871b8d55a4e3bcc2e5f08ee2d3ff1b22f0ddb,cd79a0695612cf2973e2f40d49fcb7c5d19563d2fb720ef69cb9aeddbf5bdc6a,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,74.17078571428571,74.17,exact,canonical-source,profile_27322121cfe7d870e70b359b,reported,unknown,,unknown,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,ad5a5f2007ee39c3de3855146704d005e804b1cd8a340ae8af18b48ea7e082d2,b35e205fb2026d0fe75b0db6e96838ec878dbadaecd2711a24b816baae632fce,4589a852099f2d691fcd122c2a19219befde1963631538e6595ceb9eab750ee9,release_1f4bfa5486eb4deec4942449,HLE original,provider-report-no-tools,accuracy,percent,true,37.8,37.8,other-supplemental,supplemental-source,profile_bf407507bedf02addf42d3d4,reported,unknown,,unknown,,false,2026-07-31,fresh,93,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,deepseek-v4-pro-card,DeepSeek · V4 Pro 0813 Model Card,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,d804250c713f3ec314bc053e5a33752ade15e7248cd2be9682a55ebca5d563ac,d0024344ce46cac3c345b8e2c5d95f2754da8a6028258cc1ca400b32dd6e1b7e,db5b4c9bc368ca3269fc65cd7577a445e70e6dca8d81c2798448ac5be6d6d1a3,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,38.554217,38.6,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,simpleqa-verified,SimpleQA Verified,factuality,9fd59210f4d72af3529fdd88b4e6775ba6858be12c97d96f9ab43e2de921b310,8b5fe27f27da7d15243e3917e33e94c7db0f2b1b31b5043b3c18544aa86e9445,c508a328505e9ef3dc1b69c8bfd1933ded821747cd869563855260a91233d886,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,33.633633633633636,33.6,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-08-27T19:30:16.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,gpqa-diamond,GPQA Diamond,science,6e52a8659e46b93a7e1bbb3006dc36fbc5214e9b79ab2452fb92964449dfc026,f247ff16a54f242fd02367b8130a0a7f10590d8a211a32372106278e752f036f,a493bf3e107ef04ea235ac47fd6425df7effa7e0fcca8088c1464d5bfb7ccc0b,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,89.899,89.90,exact,canonical-source,profile_e2fefa2f41df7df75dc708d0,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,mmlu-pro,MMLU-Pro,knowledge,4f72742f5273877f13daa672745c1a08172f20486e6748028f27728af2f56365,a3b8d54037a93d87cf50c3d753ca1169d4490d60b3da5ce40afce26c8cb5ce28,e3688f6717b481e55df0b8d913033ad3a42d966e1043cc47d44f24db1f27b9ce,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.206,86.2,exact,canonical-source,profile_e2fefa2f41df7df75dc708d0,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,8ef87096a5eadafa6e0c6796826b219794f9acce4112ace9afe588c31cb41db5,5c26b3216153b5b80ef1eede0889cacfa872d338f1e3b08e0980ca308963586d,48339610facb782b0d911290d11ed2dbf3217eae2adf9f0f02530b418cdbab4f,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,94.44444444444444,94.4,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-08-02T00:22:31.000Z,fresh,91,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,frontiermath,FrontierMath T1–3,mathematics,6147967342e06f6efc18642b380246f7c64b17db719b7252aa3ad3f096117127,bbba82ef692b117e4a9f19d5030c3e3ea0e638fd8f466e6bc433c6851ac81832,f2ceab1ba37fc2104982d325b74da10d78a9ead35ee022f27dc7cc583476e02f,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,57.54385964912281,57.5,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-08-02T02:15:37.000Z,fresh,91,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,arc-agi-2,ARC-AGI-2,abstraction,0c8685db78ca3dd1d3cfffaf5ce8c0ec355303c49d132ed9133cba7cb6cdb952,74debfb575427a3eaa688f93224e07b159b4d6bc96e0dffd0e7cf7be32ff6781,ed6c05b974d18e2d530771f8572131d611f108e3703feaf748158ccf388ca042,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,61.38888888888888,61,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,frontiermath-tier-4,FrontierMath T4,mathematics,7fb43cbfa887d1bb0272e16ed11c9dd2e14c4f641743f354b75d100bf369b8b8,99d736ec99deed87f7e56a6089a6b209dea1d1ce4b889c26eef385eb288dcb95,b6075d68f0e8d230cdba359167d52f612e596f30c0039c7dab990949d63c86e8,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,24.390243902439025,24.4,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-08-02T02:15:37.000Z,fresh,91,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,critpt,CritPt,science,a3cefee15eae5cf386fe0a356bf757f7049a05a9a15fd369c3c65bf41f52e3f1,df7bd7ec1d910c0278d434bd48d907eac32bd852087f6b1a77e0cc0312307cc0,71c1ef517fae994939ed6031a7be314280d251cf3237e2d39146cf104a88969a,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,16.571429,16.6,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,simplebench,SimpleBench,reasoning,56578a8a66c7d249d1980a2521c5276157479732f3cf37342e9b0d8623dcddf4,4a8d973bdf5968954d76efd76b9e1527720c2f686dc191eee46268189548fcaa,cf7f78b05898922292516882047806d7f9daa066f7ec01a61f921883183837e4,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,61.1,61.1,exact,canonical-source,profile_e2ef717a623eeb73e7398005,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,chess-puzzles,Chess Puzzles,games,fc03eeab91a73237aa901e3c73ed6bd913d311d8ab9cef4d69c4c873ce960344,4e82d51872e65fbc2c5daaf3416bc8ad72fbc8912d2e31f5415623c410c164b8,2f9516114072525c21b6dfdaa314f40c5cd10e98e8316cc20e516721c79868cc,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,32.0,32.0,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-10-13T13:32:51.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,arc-agi-1,ARC-AGI-1,abstraction,542b43a6fc992fca0aa185d559baffb6c405eb07cd1a5d5db4262d364c3ad446,1fb24d524188d5570b02719ab1fbc8bbd7cc23fb04e03fa34caed54aa51c187d,b4875e94d18b050b87239fa17e624b0ae70852c2826cbe5d3cc85ecb71209dcb,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,89.0,89.0,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,swe-bench-verified,SWE-bench Verified,coding-agent,c5dda6ace1e0d6c0cd985e3d7f0e0c1bc35711692370dab042608c20c147401e,ea7c2c9317f93576459bdf7dc7e4835716489a3107b92b980cb3c2ef751e851a,7e2388441b97b254d55da72440e4cbb7f9f784aff791ffef0c69af721401d1ae,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,88.8,88.8,exact,canonical-source,profile_e2fefa2f41df7df75dc708d0,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,frontiercode,FrontierCode 1.1 · Main,coding-agent,36472679e6e1cc48c1c662ef5eb42f71c717de51e2feaf41ff2ceddc543aff70,e10c1e33011dfb1fb725330e6bfb908992888d74638cf7e65c94c1bb8a456d23,62c5b8fe50324737434098580766a1a669322fdeda2b5bd427690c7dda11acfd,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,18.8,18.8,exact,canonical-source,profile_e2fefa2f41df7df75dc708d0,high,high,true,fixed_effort,,false,2026-10-28T21:25:01Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,ioi,IOI,coding,47dd90691b404dda755512ed1e52ecaf72a467705eda854647854fce6536c021,45ef8f9eb9e51dc8bd633bf1542600b73ba695b5be76821e52fe34b76a5a92dc,e95e16071e41f3f6ff236aa328f0bff0fc293f21b402d4d26fc3744a6ddde8a8,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,32.722,32.7,exact,canonical-source,profile_e2fefa2f41df7df75dc708d0,high,high,true,fixed_effort,,false,2026-09-14T23:34:42Z,fresh,47,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,livecodebench,LiveCodeBench,coding,0e099231e07a376fe7f54551b4f584595b396edc9e36174de182f4ef98149ba9,3f8bbd9c65d4dbdc105b39174b811be7bcb22cf0c118f133f76a22606129f8c7,d08950f44f333805ad987964e015f27b137f72ef2adeed192bb9cef4aee4c519,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,87.264,87.3,exact,canonical-source,profile_e2fefa2f41df7df75dc708d0,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,scicode,SciCode,coding,91e053a8e9c888684e9f04d65df85cafa2f977af813f17fc54be04004d9bdbd6,a48e4e012a6bd5a08471dd80f295282eb530e9cb4e5bd283f59a0264e5fa7d60,13c6184b21680ba605fc94a5455bcdeb451e82d3a9e2691e8a8aa4708d51f2ba,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,49.88425925925931,49.9,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,ale-bench,ALE-Bench,coding,a495c326385cf0f59a02cf387b3c4255ef989c6d46a17d5ab00825785ae2ee53,c8573542bb48f370111f84859594cca43827dd3b81849bfdfe3bc8316bd3ee8b,a6ed17a92b410aab877736161aa7f4ef805d5fca3b89423114f840c6c590c7fd,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1306.08,1306.08,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,vibe-code-bench,Vibe Code Bench,coding-agent,8df01860ca991546102385163c2a94ffb6a95fdaff72a03fbdb0180dcba25e43,839dbdeec15daf2bbb1dbf24b37fc5d43a3b110607b38e4f47d68ea23fe4b1dc,c2933e9fd40db6e389dc34059f2bed96f4c6dce396d59f3fca4658cb0ea014f5,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,74.736,74.7,exact,canonical-source,profile_e2fefa2f41df7df75dc708d0,high,high,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,webdev-arena,Text Arena · Coding,coding,47ea0ee120aaceb273484ecb3d2fa48714dc5770c0ba2299a38eea0c74db744b,e24adb6f96bcec18c8839c0d79c260cebeea1fc688f24391b1890e017d7d09b3,fe5398199e5b4ae0f55fab2a4bccdec2bf4197cd7613709660eebe0894a7cc51,release_65fc140e7ffbb7bf62021f84,Text Arena Coding current pool,text-arena-coding-bradley-terry,bradley_terry_rating,points,true,1580.2,1580.20,exact,canonical-source,profile_e2fefa2f41df7df75dc708d0,high,high,true,fixed_effort,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-webdev-arena,Epoch AI mirror · Text Arena (Coding),https://epoch.ai/benchmarks/webdev-arena
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,programbench,ProgramBench,coding-agent,d8082f046ba3ce8b7f6a06a2013fec65cd5e01843452e6bb4342206dd651cbb6,91cf71108cde370876542c7dccd45bc8623b20adaa9b446a3012c75e4a8adb52,11c716a896d79fbae3a802bb0201497d29704bb1466e74a8e715a17357dfef99,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_e2fefa2f41df7df75dc708d0,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,9ba6727dfbe84818e8175601705078c3bec47d6e2d04a5fec32c0de725e86b77,4e935a4e1e740cdf39fff0d487c5bf0f10d8cdff36583309b86f236132d91de8,350e65a26f9521aeab9a9f711fe352e059578ccf18a17854574c54c586e637f6,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,67.041,67.0,exact,canonical-source,profile_7379d89dcae12f3817cbf811,reported,unknown,,unknown,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,b0e5da1d8ede094a20f0d9815cbcd945817a77a569b7596314bae7f26c637723,821ed1ddc4559a9fa614432cfe5b6f2735fa1468f7f55c44ab6df20de13b59c5,3ec3dacfe32d92f6c5c22ce1ebfa3aab44f1b3892d0efa4cd510970e2aa31e6f,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,18.687,18.7,exact,canonical-source,profile_e2fefa2f41df7df75dc708d0,high,high,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,c2f227e5f33bdc6f8d1c6de63ae44fe8ea52cccf7ec4c763b7dc396a732af192,124a3ce37f95f81e602245ea38783425171cab20e1c67242d6966eb668974357,fa30a46d3c4e2647076ffb19dfd97ddfbc7fbf8e0338826a7b14058b3ca2cdca,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,39.381443,39.4,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,6464faa4e2fbebca3e5182f7093fd0c6945e13748183bfc307fd861f43ff98eb,4d6ece8edb99ec063a9e6f05548d6cb07bd48cfb9fe2d6eea1f123b1a1436a0d,592e7900639074ed6bcfbc2081e47b1a4a5b3542ed0f3fd1ad18d6aae531bb97,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1440.57,1441,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,f70947e52f6167541396653c88ba451cee9be0411c84615c8d13f01f12d991ec,c27dedd92a2080121dbf345ca8e37e665ec90be9cb28d5d7507a0c816412db5a,7e7b7e936a217676774cd60e29e30e7c07749c22aa26b8a2a3e7d145a59d27a1,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,4.286,4.3,exact,canonical-source,profile_e2fefa2f41df7df75dc708d0,high,high,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
deepseek/deepseek-v4-flash-0731,DeepSeek V4 Flash (2026-07-31),DeepSeek,open,41,weirdml,WeirdML,coding,0e35aea58b9299e908907659d4566cfd050a4f38d875b17cee8922658d7c1705,94864af600a493830bc78a5c6264917766c2ff59040709dd4328902b58333ef0,b28eccd1504c39139f2e046ff9f2e5210909468615c499a8547165b56f927d6d,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,62.96000000000001,63.0,exact,canonical-source,profile_58929078b91d4914279ad26e,max,max,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,epoch-eci,Epoch Capabilities Index,capability-index,d3d40fc5ae9707366d1aebbf94bc2b10029fa1b0a6377e59bdb9299561f0f536,8e87dc8a22acf0ecbf76ebc9d2725a34b8de18e17ad54d56d26196433d9c560e,4b56be829d24b33a8a4e6529f03f2326a6fb53dc965554b5984b604e1017bb73,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,154.69,154.69,exact,canonical-source,profile_873f73507674253f1bc4bf8b,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,arena-text,Arena Text,preference,18c1f92dc657fdbb58a01cfce3a77a68ece4ffad37f9a5d33d9b38926ee2cf50,b3a6561993ae97c9cf3cfd4262f3da51108e3749fbbd4228b14a128faf6f75a1,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1483.155122800798,1483,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,livebench,LiveBench,general,5bd0dc44a1f02cad43c6370ea812a00250ddabea1f41f411ee1dfb439c6b93f7,dccca5bf6a407e42fa846c36abd0fd6901ebccced285ccc1fb331c65ad863f36,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,73.5875119047619,73.59,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,690874c0cc7f90adad51306fcf7c488a7f0f2b64f75bc276ef0baa7f89c67e4b,ee5de3c7354b1d9f8132e037f61f29b0e3ab40ebe7efec31f6c0068cf9867c7d,d3ce1e90a3ea85678481bfab2cdb77c323a79c154bc3d4580427b79aededbf6a,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,40.824838,40.8,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,simpleqa-verified,SimpleQA Verified,factuality,fe2c28fbd1d96a7cb472595cb4daae897749d206df8a14a862bbdbb2cea27eb3,c46de95e158ecab2b75ae8ed17e1c9c645cd8ea7a19e5fe2b9a5e7826aa3f110,0d6ad28d9fc396e46a42355e03f980b920d813e637e068660351d9f66798b9c6,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,66.2,66.2,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-08-27T19:30:09.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,gpqa-diamond,GPQA Diamond,science,e15a4c232e728b0187c51a895c1e8d865495f179410a75aacfbad79bd1cda926,8fd5b86f9aaee7f343ff4df1817627ae1847006d12ee57d94e9ae76d2eac93f3,036d248110c1387eb142bde1055c6c986220c4d124dc62dd20e0c536c5e56cdf,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,93.434,93.43,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,mmlu-pro,MMLU-Pro,knowledge,543e40cde08078508afd64b49ec9b0f3368e5057c24fce9908612856b2fe5d78,d272f7cd0596e489e4ccf5f648e28292deda8ab16323f0672960c61a3577d77d,275a021c7c1f6a5743dd1c0e531db1725777e55f7ac5eb359c2479240f867131,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,89.276,89.3,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,19722ab6b659ac05e97cafd18cc2a093f68d09d60e1945f87ff50aa15c6f66d4,44c7568bca5460823d77ebe9d88bf63405364486db9bae9287f242dd695972af,4f13d3d4ab35bb573e108a1e14ddd9d6029fdc2f9388cfd160170901439a0866,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,94.16666666666667,94.2,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-08-02T00:21:27.000Z,fresh,91,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,frontiermath,FrontierMath T1–3,mathematics,3f7df3de999b798a6761bea811826438d8037d25054875aa8f22bc29a0ab6b74,4e3a5916b3615e3556e77773b1dc66451824c7a93d3730492e66f7bcdc8aa8d7,7009711d176684ec2286b418347a8c55d59be076e6ce742303dbfab48fa05624,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,58.94736842105262,58.9,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-08-02T00:36:45.000Z,fresh,91,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,arc-agi-2,ARC-AGI-2,abstraction,dcc3c22f9abca9796a3b910d6fa0af3a461b63a2d5209abea6f09767834f306d,b88a77307fe60987d62c9e9653aec287bceaa2898e472196e65faaea951f91db,554223a6af1805cd60039a3dbf8b91eca97034772aae1bb886053a411bb9daf0,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,60.416666666666664,60,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,frontiermath-tier-4,FrontierMath T4,mathematics,e6f821659da220ec55cba3f836133ea6ebef8c4c8d43596a29a654c188220a35,e1f8d9edb3f6376dbb9d924cb56ba2bdb43a6cc3f912fd1ed51003793a3cacc6,7157a5167fbcc14c246da0a6790e6feb872cf05de840639b974c69b75ce336f0,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,21.951219512195124,22.0,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-08-02T00:36:45.000Z,fresh,91,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,mmmu-pro,MMMU-Pro,multimodal,aec7c78d88488c2d8a43a8a02869b285ba005bfec3a10a79449774856e5530b1,3bee005638e10f80df59b17622a00dd0dd475eef51b66d3b772f63067f415cac,cd10394f6643ac9d38ed8ae0bd6e88b3a2dabd69ad8644b6eabf05c35b2530f9,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,88.382,88.4,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,critpt,CritPt,science,7a556887a45287d04b2c4e338c583daa7f59a1b1f8120a3b22ed96814505f182,fcca66981b3def7c7877701bad8fb46fc198dd8c9ebd1e05eba97bad7606c9bc,4c871a6888ba4d35f9162884c2d8489dda58047701c7d3bac7440123d696c17a,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,10.5714285714286,10.6,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,chess-puzzles,Chess Puzzles,games,d134d5270a1803734a5fd8bc82f6e24eb9ad0cea01da492fd51afbb52350a910,174d2e452266833eea6765a07cfaa33da9b91b751b287fcb3d214b65255c98de,eb47023cda6163348f078a26f601763f40dae842b304a97005b33ee5388ac261,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,40.0,40.0,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-10-13T13:14:52.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,arc-agi-1,ARC-AGI-1,abstraction,5cd2f302288080a63a3185c8fac88409fda121f474f49bc32d01465ec2c62f1d,8d277eaa40631697c6ed812825149854476347c08613bc56f226390eaf31991e,0ff0cecd15fa0dc7c17389bbe906b92d29f43f4496ab963d0a22a57c92b920bd,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,91.16666666666666,91.2,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,swe-bench-verified,SWE-bench Verified,coding-agent,2bdf07bc35a1678b96b376b422c421ccaaedc0552353cf16142ae887e0086031,a1642b35949ca01c9cf0b723a4b50f6c5db706dadd153b2644c7430dfaa499a5,b343a65e31729dc33a50a6f3429b8b2cc86fd1358b12fb935ed8e738b4d8b3eb,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,79.6,79.6,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,deepswe,DeepSWE,coding-agent,5bbf6f0bb61d9ec86f5d65c6e22d3a4e43edbb202c5be6ca50bca6ccd65262bd,3c5c812d091e53f9b1392a08c0d38d56b4b55516ef8be2abafd11ff6c82052ae,808897a2ada0f4bd1c0907516eeb1e63004ffe30d1bd8bbfb33ef7e9bf5968c2,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,46.68141592920354,46.7,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,frontiercode,FrontierCode 1.1 · Main,coding-agent,263b670fff3eff0d63dc5636a3908326f964bfef123f6aee0c924127f140ba33,eadcb0685ebcb389a3f80f5c0bcda69b917e3d1054b78ae96844b69a16184cf7,04560db7b711e12dca40285fbda63eddf2e4cef9dd4c3f7b4dc6841b47aa2bb3,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,34.37,34.4,exact,canonical-source,profile_d6f7288969ac5442efd6b4b1,medium,medium,true,fixed_effort,,false,2026-10-28T21:25:01Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,ioi,IOI,coding,fe646b488bbb5255a3df4d00cb4b17bf24d0f8a7241812f70220dcb91e58c576,cfdaacb95a62cc307d00d3f686463b8a44aa23af58493e7c9304686292e1b6d6,521f86923c1b5498e797d1ada569c87d4c6d33b68676e4d9da9bf67adf682356,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,35.056,35.1,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-10-21T19:24:33Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,livecodebench,LiveCodeBench,coding,801dac345091934fe2f8abcfd5dfec22589c095e19b284c170573a226365d578,15385c5f4d2d43f1e7df7d57cbcc4dfc44ebeabec31467b0ea59e339a4297162,ca3253b1789b642d754b915e20a1fbd84b6fedbe1ea9e9034faa9f873f8febd0,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,88.08,88.1,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,scicode,SciCode,coding,22c5d4cc8beeb379d43923d4fe035e7ca2c555ed4b7e7e92050632e88aeb6235,dd92627b60a10cfa605855072dbe44cf922e11eb1da860e6b8eee30310ba0b9f,1afcc49b7d3e3593a273e5b654c5b1406c66e611a6d441ae31280b76d079b24f,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,53.3564814814815,53.4,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,ale-bench,ALE-Bench,coding,8958c13d4cdcba65026a4f9b3ecb4c56af165190b8309c5eb761e4c69c545b60,bf3c900f96f84fd763cfa5758df7774c772853475fce58b155ae43aa9664dbf6,7cd0014ca862437fbdce8d22a81e5a4e5e55089d061b1b3238417a012545d34d,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,715.52,715.52,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,vibe-code-bench,Vibe Code Bench,coding-agent,d15dd73638c79e8e205e04a74037c5f300288712fbff5e3d39f00bff8a1027bc,746afbc1fb5226f1e2f4a8b1cbe49b236714d5f914edb14917768640fc070e60,bb352c62009cec5eb45a26c704bdc44b0f278ff94621451998d7bc628397a6e9,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,64.005,64.0,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,webdev-arena,Text Arena · Coding,coding,8273e59652e0e64ab929b6d2ee0f9d21d56f9cef02a4e5f678c2853fabf27680,71980432db50c41cd891fe810ed34433472e304fb151cc4ae50f44045a2d7cb9,969a4db70abbe06850e6b2d8d7122987525372ff4d84e100761a7e6e07c01ccd,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1537.5456409227427,1538,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,programbench,ProgramBench,coding-agent,39fd7ac8a960813659c1bd3523aa3a4ee04f2520e589d25ca151146df852e4c7,e5294fc8471b84efb61832f139c10b812e4ea9674ef07020d04fcb528b91dcc5,575465df6e3319d9b9a0b4cd3562b0145a7e88ff202a93e1bfac9835326c0ea0,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,95d17feef5e944a2a21461c871dd0f3a591f9f6da19780be1a7a518b088f29c2,a1bbe20cac011e1160ef4dd1f7e916ef0fbeb493ddc8f455f5ee7564ee66f54e,6d9f534def2a440d1dbd6fb87b11b6a2d4d7b30bb4a9894c3fe2353d7518186f,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,73.783,73.8,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,ff4639c45250be2ff1be83c22048211149f3c56cd3a39aa013834ffe57b92d98,5fcb31d401a9479d755bdace4f2c6770ea2bdac2faedb91bbe54400ce376ec64,512734a6a28523bf4572c6c83b3ed7007e69ec3a01528bf06a4fc4daa040feef,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,29.896907,29.9,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,apex-agents,APEX-Agents,tool-use,1b05312ab043561c02314b545f8fb963c8912d88d4790b1025878f1558d227c9,f7735921b21f326e22a55bac145fb07fca7f44e18cf42a1775e723ed6f27c919,4fd14524089ba16d3f3fb9ca6d6097a22820734676e76806a473ab8424621689,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,46.9,46.9,exact,canonical-source,profile_8c08c2576548098891e1714e,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,2b8fc35112da238a4cff83e22f32ef64a0fee4a6ae3741d9499576177c1e11b7,f947b27de3395686ff705961908f1326e7905f69a51a63b156749d275ec2c7da,329045579771cb2f4f33fd76ab1dc151844151e977b5c771055765ba824c8527,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1294.63,1295,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,5c53392b93820d29727925bec2c59a9c6fa1341a3044cc41181f7c400593d8c4,eb431d120e0e8e93ff69bebb8c9fa8adcec090e5fb087534552e2dd9f3cefd2a,18ec0db34175ebbe4736138c30f4399c24503d43420c8ef925e43ef33e6b2fe3,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,4.286,4.3,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-10-26T22:05:28Z,fresh,5,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
google/gemini-3-6-flash,Gemini 3.6 Flash,Google,closed,42,weirdml,WeirdML,coding,103de9a2b89c02a59e077199060a7d67750e912156b5f81ce6e82a61fd2a8876,c9ac21ac3075ebee6ce28d564c03dfaad8b40dd47869f1e0bdba32b487184548,fed35aefbb9ecf5bdac93fe42a07a1389116668de3e39f8adb4adf526dfbe273,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,56.10000000000001,56.1,exact,canonical-source,profile_b910cf792225dccb483afa4f,high,high,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
xai/grok-4-7,Grok 4.7,xAI,closed,43,epoch-eci,Epoch Capabilities Index,capability-index,5bad1561b93ddb3033df7783f4838eb34147d4cc4822e48658d4f9f0f4afd6c0,0bd89639cd85189e4f5bd9b328892dd77cd20ee09179d900e756a690315e2d89,2cc6410bec453b49e8e33a4abf0b045c103dba0883f0653799c5f170e42ce805,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,154.55,154.55,exact,canonical-source,profile_41d3dbb96d2673ac6d6ba6bb,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
xai/grok-4-7,Grok 4.7,xAI,closed,43,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,44c9bdee63e731eccf4c4f0d02f72923ce0b1bd9e8213a5b10a2a0df8684df55,a05e83a4de4c0818c0d2ef2ef953593d0c8e90533744a4bbf713aa907a0b2cd8,0a2531931526defad170f29b6712fcbaeb592dab72ba66e013c91f9d49eeaef0,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,34.335012,34.3,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xai/grok-4-7,Grok 4.7,xAI,closed,43,arena-text,Arena Text,preference,4a6e2f0c00344631d517c5d0956e193feedb18643f4bf564d4e5d7f01acb84c5,dcfb7374f3fd9e1e2ee4049198564915b5ef3ccc326611aa9b023ba07f5a5ee0,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1441.2270542052997,1441,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
xai/grok-4-7,Grok 4.7,xAI,closed,43,livebench,LiveBench,general,77c6fe57f4ca3e4d1da1870e6a97ac26753aef3dc3fab5d915f2f2ddfcf55736,e36b7edf5132d478ef50ec8da3be1fafc8910550c4a018f9ab48504baf38e0b8,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,77.39673809523809,77.40,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-09-30T18:53:19Z,fresh,31,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
xai/grok-4-7,Grok 4.7,xAI,closed,43,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,24be2aaf9baf309481c1318d8297536c3fafee42dfa5c19e946f68abb72265d8,7aecd7a7f8517f518150c7d9cb38ee8c4dd079f3e63fd3ea3280b725a436ee17,09baaf16588d7f5bc1e8de8012f98dab5ae18981e2a53a7c6e4436443eb3a964,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,43.141798,43.1,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xai/grok-4-7,Grok 4.7,xAI,closed,43,simpleqa-verified,SimpleQA Verified,factuality,aac34a5d4373f0d7414478d1012f33a6baccd126c99a95533bfb9cbfd3814154,07ba013d843a0fa5cf903364bbbc925601a36d4a560cfe61c97759bcd18ba6a8,43c130d8c464e692585f692717ab455f3f31e03573c611e2a24208f0bded20af,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,56.00000000000001,56.0,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-09-22T21:49:22.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
xai/grok-4-7,Grok 4.7,xAI,closed,43,gpqa-diamond,GPQA Diamond,science,b090e548cf1b6c6225b69dab169bf0cdc202f545e72590d7848591835860eddb,e216f0deea3261041ed83097fa8bc3f5515d6a1758b959b33b37f60420773ba5,c7166913d39cf25ff6883b56b97c6c4e2c2069c08eb39b06111cc2a769676565,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,92.73989898989899,92.7,other-supplemental,supplemental-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-09-22T21:49:22.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
xai/grok-4-7,Grok 4.7,xAI,closed,43,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,1448e36cc7faabe0d0b509b012f4d753ccb657cfe7b49f96ddd447b5d617ad79,41aba1ff096d74786d55134b1164790cb1069a372e52b938a42d00a799d384db,3b87eef345d7e339a4cf86dd308fb014be73564b5beafbb401d4930cd7ab9441,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,98.05555555555556,98.1,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-09-22T21:49:22.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
xai/grok-4-7,Grok 4.7,xAI,closed,43,frontiermath,FrontierMath T1–3,mathematics,cefd0f96500d72ee527bddea73b67f2415f75602e17d24cd5aea1e53a1adbc14,05c767def89b7c1b083f49060c09e6fd8dcd89b881caf80b6eacb94f19612294,ad64bf406a59155c8c66d3acacd28cb226ec54a10ee69ae99d3b1a4369969769,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,52.98245614035088,53.0,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-09-22T21:49:22.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
xai/grok-4-7,Grok 4.7,xAI,closed,43,arc-agi-2,ARC-AGI-2,abstraction,988c91df41c2bfa727dbc783e6f1846007b2963352e78c09f0ae570e912b8e5c,4e6fd632725586911fe11948ca1ccc21085d4058b9a58efde5378a673aacc7c0,ed4a2a4a139d50926900b1501b97cb17353f657136e72bcfc3ea78973dfad2c3,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,61.38888888888888,61,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-08T22:23:19Z,fresh,23,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
xai/grok-4-7,Grok 4.7,xAI,closed,43,frontiermath-tier-4,FrontierMath T4,mathematics,1fc80fe347264ed65d0daa28957bafe9f1458686b160af8ee53ada2e024ca98b,96c872fb288b9628f0cac265fb62d25721393a699fe41d265ce5853ec2f54e93,8f425196ee2bffb2b6982024f14ea775092e191848a81f4c192683d2b8aeee52,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,17.073170731707318,17.1,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-09-22T21:49:22.000Z,fresh,39,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
xai/grok-4-7,Grok 4.7,xAI,closed,43,mmmu-pro,MMMU-Pro,multimodal,f41329b8edf8d075e5038479d62c4f93c0b40c37ab9bc7dc3e77bd2596cd070e,abe4bd6c4d2ac286f1b20a7517ff050ac47b3421700422d8a306e63a0ddbd221,d7231f31960e21887048beffb2f74e66289b104736222c734247cea636b1a19d,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,78.150289,78.2,other-supplemental,supplemental-source,profile_cc4e54642b60580b7257387a,low,low,true,fixed_effort,,false,2026-10-01T22:31:28Z,fresh,30,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xai/grok-4-7,Grok 4.7,xAI,closed,43,critpt,CritPt,science,7dc9cc043ad29fbdadc87ad4991091312cbbf2ac8758663f9559a41e9e38ca01,40412432f45a51f7157eb75b6f71133541d4ca9e85e912310522f0a3af884908,dd8f41a4d230a885e45d7e16de85b502999fd96f95a7c0032d8ef6cbd69e1df1,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,17.714285714285698,17.7,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
xai/grok-4-7,Grok 4.7,xAI,closed,43,chess-puzzles,Chess Puzzles,games,a689870f354dbc5debcb77142fdbc4fe03c84a380cca7021e1b290b875772d4e,2f1fbe706cab72b00363d584ad001656203a3ed372734a2631003597d14fa62c,e1c987a055d9b0783be516d94254efb2f1e5500bef6d7849285983a5d1ee00ea,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,41.0,41.0,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-13T12:39:21.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
xai/grok-4-7,Grok 4.7,xAI,closed,43,arc-agi-1,ARC-AGI-1,abstraction,ff2413ccbbdd7db2d1ecba7018c9b4572064a69193bb67eab9eeaaafad67dd13,9d18bf94f2565bc15111190bd8b86c86fcfed9dbc57035d984ff54941368b172,1d72fd0e84f6ede463fc5584d2b4ce1561454cc537705fe0db3057e075521547,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,89.5,89.5,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:14Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
xai/grok-4-7,Grok 4.7,xAI,closed,43,cursorbench,CursorBench,coding-agent,e8acd65262ddfb582cce5c930a55c717ba2371fb5452c8341e8ce87ce0bf411e,31f4caac56926ec9de6e973df2b25cc2cbdd6db0d0caab14b8ae43a1b5d7c7b0,6cbb1948644b8ad8e8bb7940086710098d626a0b5447950e21313d42b08b488c,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,46.3,46.3,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
xai/grok-4-7,Grok 4.7,xAI,closed,43,frontiercode,FrontierCode 1.1 · Main,coding-agent,f41be5b34e4fbfe8ce067d48f09a6c8369f574bd676e4e0eef4bab27d3f0b614,afa141174a989803f6f895c920d403bfc71f371c77f47e06e497be1b9732cdf0,99e4b51319b947166a810995ac4bc70f908c5af0f32398ff03a4b015b7d49cf7,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,47.589999999999996,47.6,exact,canonical-source,profile_b1f23152c5d73c89b843de81,high,high,true,fixed_effort,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
xai/grok-4-7,Grok 4.7,xAI,closed,43,ioi,IOI,coding,6aba23d8edd08e46344f9582df9ab0391de2a090bd97fed0a83f8f666ad6f9b4,a81e4cf6a02455bafe2c8801b1b11f1aa79666dd02133aa16cecb1c6945f2b5a,e91eb17b8aabb913412d0a64c97a778c7d5f39fc41357f709c9d3521d046e2a5,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,57.722,57.7,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-09-21,fresh,41,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
xai/grok-4-7,Grok 4.7,xAI,closed,43,scicode,SciCode,coding,47d560278655e6452d2002905118eb3da8134c3648f294d0c9f727ae6c209489,5068f0a09de5cfbc0fc1fc2dc4c80388e8e20ec64d3c91d781dd74c9e7009b3f,a076f02a0a1cabf30ff29d1de05cfc8638bcfb219418daffee8ea3ca5f8ee1ea,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,57.4074074074074,57.4,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
xai/grok-4-7,Grok 4.7,xAI,closed,43,vibe-code-bench,Vibe Code Bench,coding-agent,fe271c4c856592da2451fab5d1188fa69683b395cbf28e81c9db0759d687ca12,5889713918388e329e60a4b2948c65c014732c771b1c786a1420088e12668f56,da3ab17f506dcb1435a5be4fc1dfdc76860308e58a0887ec74a1983cda8e6c6d,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,86.175,86.2,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-09-30T18:53:22Z,fresh,31,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
xai/grok-4-7,Grok 4.7,xAI,closed,43,webdev-arena,Text Arena · Coding,coding,c30cbb090dd69473c8b39f0ca2a9b809213f5974bbd4eec0e0aa5ed05d1b1b9b,f180c8d09d884e6850d54eb9a9ec2ee88c5f73aa309ed9849e03bcee713204dd,d0eef5b3aca0d99e446b293ff5ed4773674609492f3106653f147842e994e5bd,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1637.4572057581888,1637,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
xai/grok-4-7,Grok 4.7,xAI,closed,43,programbench,ProgramBench,coding-agent,2456ef9c7a7c6d932c83fe384f21c607e0c28d20348514e364e1ee1a148d670e,8deaebecd8a549e2dc2d0cea56ad7c68b68c50ea015b9ac1d324ac425c02e46b,fefd96bf3cc9d55c2d71409e0324e0af7ddacc60a19bdabf8a5b78817cdbad31,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.5,0.5,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
xai/grok-4-7,Grok 4.7,xAI,closed,43,weirdml-v3,WeirdML v3,coding-agent,a8d0de3e87d58dec3553f0d1b7b2a008eece8b068d357f916eda9dc97ba1fc08,70c762d78bc471b71799de1c8b6f90a9f3729f6b27164fa51b00505972c9d2b8,9e4f94b8c03d7f64b491f2338d82f7228bcf29e154c1a6d90f2ef45087000b39,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,8.452663636363637,8.5,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
xai/grok-4-7,Grok 4.7,xAI,closed,43,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,1ecfd7d0148a97f138a4e76effc13e43955384dd34fa304c1f66b15900750656,cc143346cedcda08e67192268ecb7da2becd2e1a938d550d1d3ab26e141d6a8f,b9748d5e3cd8dd853b51324ec685115602dbb87a372d953b1696b8b5791c6ca5,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,73.408,73.4,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-09-30T18:53:22Z,fresh,31,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
xai/grok-4-7,Grok 4.7,xAI,closed,43,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,283d2a85dbf853ac34a4b3ce7b0500d4518bda55c264400cddd21390023a1841,69f8bdd530e28edec395753453534f0e708273a2cf23afa5e9c752a146f1e47d,398e926d9443cd1abf77adae8779775a763cb722d025bcced02d39a5c3863a35,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,28.788,28.8,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
xai/grok-4-7,Grok 4.7,xAI,closed,43,sre-bench,SRE Bench,terminal-agent,f10f0d85821a7506bc608059fb4c13932d57f1199f6a627a384b9d82b8197467,d4cdb78dc92f0cc6b34a50f1b796cc9d5c70d749794d2e9f4f4121a5e70801ec,6a6d24508a05f708213a21c176a001b4d49f3022b5c32984791443de5dcf295f,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,3.817,3.8,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
xai/grok-4-7,Grok 4.7,xAI,closed,43,apex-agents,APEX-Agents,tool-use,6b1e769eddfda919ef337c69b45235593fe0e4e6ddc4b7ca0d12e059c138abeb,72140053929002d789b6f5fe0443241afcf7a9bedca50f5748a1fd977381ac03,bbbc5bb3cc34d866861fe43d73a30e9e4d58ed7aa8888326d2f0331a8f919e51,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,54.6,54.6,exact,canonical-source,profile_89a7d6172adc99c954f9182c,reported,unknown,,unknown,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
xai/grok-4-7,Grok 4.7,xAI,closed,43,arena-agent,LMArena Agent Arena,multi-turn-agent,2a82adf52350e4603d9b588942ece65d520fb0db01f895de19abaa5ee85b34dd,c2c4d00116a2a9ee60e3216ba0cd675edcf923d25ed93d933bd58a9ae15f43da,c997f4ebd042cfcda34d7cc76ab45791850e4c0d4c9e22314656ebc8dec06778,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.023946701791754732,0.0239,exact,source-specific,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
xai/grok-4-7,Grok 4.7,xAI,closed,43,vending-bench-2,Vending-Bench 2,multi-turn-agent,61e0c7bb2e51ce788900a66c660b014ec667cca553d9b039487ee2eefc28b1ce,5e1f59ec090f2cc839c2d7c4d17c4af2b37c0a33015e9c1ade7339a9b6d80308,36f069c55af8c61cea81e24a730a10a585696877835b2f60a92ae7d9ea314372,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,10536.831666666667,10536.83,exact,canonical-source,profile_4b6b7e85727e66180cb268e6,reported,unknown,,unknown,,false,2026-10-08T22:23:14Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
xai/grok-4-7,Grok 4.7,xAI,closed,43,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,b2ccbd512953a6b94f1e013556df66f3114fb981725c4711a7eb04494e3720e4,10c3b2e5c0524628e7c338d4634df755363f2bd81860d74997f1cc1fa42c1c85,95fabd5d8840686f1fe5a5bca8d8fd618bb4372400b50f5a1fcfad3a18a39c58,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1718.42,1718,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
xai/grok-4-7,Grok 4.7,xAI,closed,43,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,8a82a587a93bccbec2db67cd17e5068e898927b833b9eb6f3f4506ce5151c9ea,c12e3eb06c3ff056eab28d1fd30ebbc6d2de7ee06120f318a6300a1503f0d98f,ffc7535fa6bcb9010312382c5ecf9b4af78aefc293056d5484eaebd56c88a9f6,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,10.0,10.0,exact,canonical-source,profile_1105264d4579e04e2c6a4c9c,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
xai/grok-4-5,Grok 4.5,xAI,closed,44,epoch-eci,Epoch Capabilities Index,capability-index,c23d0c8ead90ff0e884d3767f3b50a64380e4fc1d6a815180a788943bceb89f2,eed941188e0b795fd227665c354c1d2e88d3018c48eac07c0025f4dd9698ed96,3356d89e877509f21c7113b9241d784d738cf58adb249c1d5943aa569f22dfa6,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,154.36,154.36,exact,canonical-source,profile_cb44654bc85aaf39fe593479,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
xai/grok-4-5,Grok 4.5,xAI,closed,44,arena-text,Arena Text,preference,8380fb710a7cac0a118eaefecb751048f8cb8cc12ccf4d715e60f6b89454ef9a,fd5e527e7bae79bd21fd1a188931825f6bcad9de23cf371270316d1ce24d6add,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1465.4464199324245,1465,exact,canonical-source,profile_7c1e5a04c3bf319f34e54098,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
xai/grok-4-5,Grok 4.5,xAI,closed,44,livebench,LiveBench,general,2a736bd66e96a719635e8aec4f56f35c5d638ce557df18acde6e67989ab7c4d4,996fff90b4da941f85081c855ba5684938c9975a9674b24fe743f900c86485b8,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,75.77284523809524,75.77,exact,canonical-source,profile_4a583e289c2d0896454de86f,reported,unknown,,unknown,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
xai/grok-4-5,Grok 4.5,xAI,closed,44,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,f6adf8d716d2fbff38b132cfcc05bbc08bb1f255bf1cc4fdfbfca6da581b41cd,34b88495ef8a7de06426180d152c4330363080c5b1888153f57d362e06ddb686,9ed249e49ef2d928fe3f70ad57eb328a3ff34b2b25fe8281440621d9c32f4087,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,42.678406,42.7,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xai/grok-4-5,Grok 4.5,xAI,closed,44,simpleqa-verified,SimpleQA Verified,factuality,31f9197b29055842baf3b68ae89d6ab91a34440a596181540da52960b3385fdc,6505dbe0444b4dc33b9696e49a581e4e3fda81961d30825d3a97f6a2d2dc950e,708704be34be87713cdc078eb35d334a987d1955cb120dd9aa7ecf74278c21a9,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,48.3,48.3,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-08-27T19:29:30.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
xai/grok-4-5,Grok 4.5,xAI,closed,44,gpqa-diamond,GPQA Diamond,science,ef9e17ed59fa4409a51f4f2404018288c2356305ba315e6fbbf57c12dac3d754,83159f4b1bd283f2d2b8c603b455d887524c1a14ab790488c9cb8da852145adc,2a8aaee35acf7ffcc1c4b7f3c771197affac6a9935489acd67fa9d5e46824496,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,92.929,92.93,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
xai/grok-4-5,Grok 4.5,xAI,closed,44,mmlu-pro,MMLU-Pro,knowledge,82d9e5855b57ee585e0bcdfa0095056c95629115fcee9e097dabc47e535a84e1,f61c65a2db448df13a57d0533e6b8fab4097a41c56e5345cd70e987dfc5f3570,30783e9e3e504467b7393749b69de46b5ab86c389a05e1c12e5ed4c6204aef96,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,89.216,89.2,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
xai/grok-4-5,Grok 4.5,xAI,closed,44,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,09865fb30b11876232e160140966492084bd57f18c50486bae0ca2a2407bb6fb,fc7dbbc9874b78d984db2db3a8889bccbdf004d03d103d2bf3c3667f4ea094e0,7cbf38020bd69bd48d8e8fe29843f017d6e1e5a9a361c88d9b00f8b0f310957f,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,97.77777777777777,97.8,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-07-08T23:54:22.000Z,fresh,115,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
xai/grok-4-5,Grok 4.5,xAI,closed,44,frontiermath,FrontierMath T1–3,mathematics,cecc05e319492c41f5b5cc784471f244dcb023783eb7a5d8a8105e0467442d3d,9e2c46ddba31a61086ad167338677381bcf3981f298c1842ae7305e0aeeca29d,b6614fa95b24c1d92a5abd67890f0e700c4381fe4735c9c1acc943007f927de0,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,57.19298245614035,57.2,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-07-09T01:53:42.000Z,fresh,115,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
xai/grok-4-5,Grok 4.5,xAI,closed,44,arc-agi-2,ARC-AGI-2,abstraction,a3cec82218ad710b7638b97c2e3ff64f5e88b0449544736d1e7a59e6b4feee1b,b30b1ddb9425bc4dc2f753fb7f88d326aaaa25f12cd213d8d4bd4f98773bfd65,71362c525a715c2754a7cff2c2545649c1c0aff25800e6118476191022c6b207,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,52.64,53,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
xai/grok-4-5,Grok 4.5,xAI,closed,44,frontiermath-tier-4,FrontierMath T4,mathematics,69b9b160c4882329d88a688d7a8d2e3f8365445c4eaf32ac5c39145c167363e6,bea3edf9d0610afa29b0bf649411d364f2a648c8a16e89f73c420cae9785e63a,0cd213659d79ae2397b7ff1939e91844219063f5993ae240a46e4183d90e0845,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,24.390243902439025,24.4,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-07-09T01:53:42.000Z,fresh,115,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
xai/grok-4-5,Grok 4.5,xAI,closed,44,mmmu-pro,MMMU-Pro,multimodal,a6df5ffcc009a231d58c987e0a65fb3db928c1eae737d0742e90b7112026374c,f32f6562fd47d849cc62dddf11ced8e795fab5ac882129eec9bb736d91044011,9cd973b156f4b2c6098b25ee60bd123eaf7b9b0ec7ab107f6adad140085ef369,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,61.792,61.8,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
xai/grok-4-5,Grok 4.5,xAI,closed,44,critpt,CritPt,science,5c96bc17ec6151470f73b3fab55b84763920fdd56327fb8cb70f720addc0b5b6,e816e19d961b770db1d8d9588b56d2664be9c2a2ec7266493d08a27b22b5c14e,01724d8acddab79cd8a6dc45e06906c7804ae19aa3f98d297f91bd3fa39c645e,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,15.4285714285714,15.4,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
xai/grok-4-5,Grok 4.5,xAI,closed,44,simplebench,SimpleBench,reasoning,0d056695c37263ff0039af3287c084fc64a1f35dbf8c3c7589108f0fa7397218,8b59d9c1732320c15478e97067263669f6c1dc4edacbd795808a68a3752e477c,7eb48decea3ecaa512c14af260f93c5830a0a514400310caae0df1b4f8cb7406,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,70.0,70.0,exact,canonical-source,profile_e2a20830119d2bc9ecee925c,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
xai/grok-4-5,Grok 4.5,xAI,closed,44,chess-puzzles,Chess Puzzles,games,12a828dff44f211c9cfdc1f056016e156fdab15ab71ccc6409f53989993d4077,bf3bb9e0512b767400620e684c253b33796c519fda3c786a127ce019d1779742,9cc88178e492b7de42f9fdd7b84ef537c11e8a58c110bdbd5c980649721be921,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,35.0,35.0,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-10-13T13:14:23.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
xai/grok-4-5,Grok 4.5,xAI,closed,44,arc-agi-1,ARC-AGI-1,abstraction,75116117690af31d02f75bb1429c7a4cc94b1495a8b4050890f41a3b5e0ba2a0,0640d6a3c7b39f42786a91cd8abc13b15bbece4ec738e7d8efd953bb7e611dff,0be15b5fd304eb56ca0f44856d57eb910ed57bfeb05acceaa66240d6bd25a138,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,85.67,85.7,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
xai/grok-4-5,Grok 4.5,xAI,closed,44,swe-bench-verified,SWE-bench Verified,coding-agent,644c3601bc5f5236d2d498666ff496e2e0f8f68b4d74b97eec604ca3afad5f23,7f246c4dcd71c90b483279b4d3f6bab453146659f03621ded6e1e1511b9e88bc,91004525e8a60922427166921e1c489608e94441bdad6fd99d92565b6cfdedf8,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,86.6,86.6,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
xai/grok-4-5,Grok 4.5,xAI,closed,44,swe-rebench-v2,SWE-rebench v2,coding-agent,e5b55272bc0c8263a108183c1c1c1c8e50c1fde3d297af19350011aceb9a4b89,fa3b83b0c0df90cc80d9e3128d37165269a9df7f39d09d19a60fd4c1d5a92287,c7ce3d3dbc15c2d5de949b75159fdd0c0d5fb476e04271616900e73893ee28cf,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,63.78378378378379,63.78,other-supplemental,supplemental-prior-window,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-08-30T11:02:42Z,fresh,62,0,A+,0,A+,7,C,potential-contamination,6,potential-contamination,swe-rebench,SWE-rebench,https://swe-rebench.com/
xai/grok-4-5,Grok 4.5,xAI,closed,44,deepswe,DeepSWE,coding-agent,03fdb0b16f77c8d43b91f1c1f844c1e2645c3f0d14f22efd0084630ddfa61b76,98c6076e0851e228353e6ec4bbb40052ddca9b08593f86502d28236cef4cfc63,92509b6905548ce6b2e9db8427bb6bde696ae1d9d7d8b138aa5e5c11b885c7f7,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,53.76106194690266,53.8,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
xai/grok-4-5,Grok 4.5,xAI,closed,44,frontiercode,FrontierCode 1.1 · Main,coding-agent,bb45dff861b188b806bd95b76ba3439d9c39cdfbf0a10d4959906bc4bdbd0477,3bc12b94bd1354cbdf79f1e3c8e8be9ccb52e8207abedd2b88f51c34b8c90cef,40f16d3459455495eb47f3033dd5b0fd57b39ce300086fbc9e2ced61045e5710,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,42.44,42.4,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
xai/grok-4-5,Grok 4.5,xAI,closed,44,ioi,IOI,coding,4f11795d6fa349bcfcac5889b44d4cbf2d20ce438d32b508ccb4be0b73aa018b,985c633390ef35d20ca3981c6e1812b8cfac14b717e4392dc574bd82f472fc38,8e196fab58e435ad88cd489105eb1c155bb2cf36433105ceb6a6671fea5c817a,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,40.556,40.6,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-09-14T23:34:42Z,fresh,47,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
xai/grok-4-5,Grok 4.5,xAI,closed,44,livecodebench,LiveCodeBench,coding,487cce8dfebe52995291057d10c071675618d00877e89229ddf0924a33c63a28,82cd7a725233cd77e7a6373d819cc0e554ca39e3ebbe1917ccc4b4cf178b39ef,3a0567c444fb0b281c0f3337ee2dc3ebdf76cfe4a87783ae18d30ce6c98c01d9,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,87.351,87.4,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
xai/grok-4-5,Grok 4.5,xAI,closed,44,scicode,SciCode,coding,937ce4456e034b5935dd1c8df5d4fc70ea7c5c969aa56be7687c907ec19467a9,bb1753266f698de6a9bb44bec517ec797a1b8bd739d1aa3190e31a7ad0ab8388,1ced07f516399abf66fa226f9eef7b480a06804ce384c488f55b773e6dbb857a,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,54.976851851851805,55.0,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
xai/grok-4-5,Grok 4.5,xAI,closed,44,ale-bench,ALE-Bench,coding,6a3d44e47a6f739815284cde192c470f64dc53fe7abc492e9743a3f3d17996df,4aa25c7c0bcdfc72fbde73196f1cdd59e4e12b753ea3bc89299f6ba7747027fb,5e43d648a673b4fb850af65f22c490376fc0f5a52012a0a59dd674757e7382f8,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1309.28,1309.28,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
xai/grok-4-5,Grok 4.5,xAI,closed,44,vibe-code-bench,Vibe Code Bench,coding-agent,d872caf51ed0b39e93e0de26da073b053524a69471af30bfcfe7235bb69dba2d,dbf52a394c9ae8e25fcee7f21d35e9157f4b35dbb3054df4c282a5b0894a531f,cbfde0619c414bd4df277224befb997e5f6578498ebb9e9643f94192843e9cc0,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,69.002,69.0,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
xai/grok-4-5,Grok 4.5,xAI,closed,44,webdev-arena,Text Arena · Coding,coding,488a681974d5a218eb9c55daf7db24445cfcc201cf85ab6c47fb97a0bce73ee3,c90aaf460373371a18fbc8e4cf73fa274a4fa2e8143f2402e60780f0b8ed8bc1,f7ef2c075889f128d0b6e2992fb0632ce71f357201cd6e6aed06bdba3b209721,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1553.032541111338,1553,exact,canonical-source,profile_0a2cd2202522f2c34abea7da,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
xai/grok-4-5,Grok 4.5,xAI,closed,44,mirrorcode,MirrorCode,coding-agent,96ad08880242ddbac427537de739cdd61c5c4ca8cfaa8a1979aed10e09de4cd0,45c9f31f1841126854104b0be42c5f29627803fae8e38291a4eb2d45f00a3d5b,290a92db8b9ecf0be0c87a2a891515e339b83acd8be75b9a325dc61e52556dd9,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,32.22222222222222,32.2,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-10-21T16:49:00.000Z,fresh,10,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
xai/grok-4-5,Grok 4.5,xAI,closed,44,programbench,ProgramBench,coding-agent,3e634d9b4a00a218d313873262457e33b3ba484989cb95f40ab5ae5dad7913f6,6b0a117f300dbdd4a58fff62e23e84be9dde5443cfa37369e1e8f856008e96c2,46edb43243e433304d7bc1caff80231e6c2094235cece27d53a400398ce45f7e,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
xai/grok-4-5,Grok 4.5,xAI,closed,44,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,56cbd9f6de12dff4a023145cb09da184d947932170a4f43d20601f7e4cb63749,17f770532af16040dd3affb6cfa9843326357bb9c6fce0e8954ee60bee792405,99c1674da391fac1fd1fbf85dea47f0c8de61ed2e6d19c0965904d9211e979c7,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,67.79,67.8,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
xai/grok-4-5,Grok 4.5,xAI,closed,44,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,59d6e460b2a909c62f8da1df66f1977bf5b8c6e55df8491598118559877f3668,5dd4a8c901f2d6972965c5a19d6790f1efd1ae0a2e7f94f5f46034344ea39bb7,a431bcc5a38cf066e0c8266838a434cfb707d81260ce4d9a6bf13c91965c05b6,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,8.586,8.6,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
xai/grok-4-5,Grok 4.5,xAI,closed,44,sre-bench,SRE Bench,terminal-agent,fe70686e9daec8d352d76562d957d124ee528823e332328fe9954d424bb35eae,311140a186bf2d7398e87e51b514daad2a09ad620b8dbd6113b09579e95a3098,db79727d67c5d372e5ec385843d53c8e2a86e4534323296a600d8a5c115f4f56,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,0.763,0.8,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
xai/grok-4-5,Grok 4.5,xAI,closed,44,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,64ba2dd131c364ca9977867113caeb8368022c7ada00028a4f9cbac8abdddf1d,dd328c124b4f720cddeb19ca8aee35e0d4fd0a4feb1d2df1d789187d7eae2182,75d8bf7afe74687378deb70dd0c118ff0b09997f4dbbda5c6f173943d12a2ea1,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,42.061856,42.1,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xai/grok-4-5,Grok 4.5,xAI,closed,44,apex-agents,APEX-Agents,tool-use,3ff9ea23bd234d06b3bb0ee3307bcdf5dd1bb3a5c955a7216fbd56dda8979fe1,b5ba9bcc823041ddeeb351200d54fc9f224b119c76d5a1b670df6f0c8a4f6bd2,6c3b64469d43afb0bb8a2917614bc92a8560dbf8c565fb26ac9de05af91ce1ae,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,56.2,56.2,exact,canonical-source,profile_7b5a51c768e676f4fd84a86a,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
xai/grok-4-5,Grok 4.5,xAI,closed,44,vending-bench-2,Vending-Bench 2,multi-turn-agent,6d4a14e7448f5d085082eb9d05ae3d96d2c7c30c69c5cafc7a46e49faaf31f65,b367e5577c4a8e8e69c31dc564a4accf642a23daabacc9db25f7264416a9dad8,a9af951f1de03bf9cc633f76e5734942b582f6a9d28975187544eeb7ee252bc5,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,3887.4259999999986,3887.43,exact,canonical-source,profile_22b572a08df9ceebd792a8c8,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
xai/grok-4-5,Grok 4.5,xAI,closed,44,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,5536839a619e4a5927a49ec7f058001b2118f396326e8bd6c3beca202fef78ff,913e3fdde55f77307a6ad75fbba7895316457cb4a7135a7add0744c34085353b,df466eda8afa1d45232539b4d8e1df2cc8fa4fdcb291394148c92305e8d6fb1e,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1389.32,1389,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
xai/grok-4-5,Grok 4.5,xAI,closed,44,terminal-bench-3,Terminal-Bench 3.0,terminal-agent,5c089362da083b33cf9a25499c6a3d560c2f76e2002cd285e9f26504335c9dee,327d9c632e04a2ae64e7b61a9765629ef0d2e0b70fed6d0254fb3c60f995a547,caad2167d6277230a98c08d572a03818e1f18c1cbb389f66f957eb8652e6e5ac,release_ceb36f25882d8cea2c9d084a,3.0 · b936386b-4afd-4ad7-aa7f-6b52bcab9853,official-74-task-leaderboard,accuracy,percent,true,15.68,15.68,exact,canonical-source,profile_cf7ce74acf6af29a63ba09f9,xhigh,xhigh,true,fixed_effort,,false,2026-08-12T21:39:34.661007+00:00,fresh,80,0,A+,0,A+,0,A+,official-leaderboard,4,official-leaderboard,terminal-bench-3,Terminal-Bench 3.0,https://ofhuhcpkvzjlejydnvyd.supabase.co/functions/v1/leaderboard-read
xai/grok-4-5,Grok 4.5,xAI,closed,44,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,531add623541a6ba7d561e60d248944d86a0553de2fe948f27cb34264ba1cdae,89acf922f06cc52fa6d098711bd3dfc117a3439012a1b61ac0380d2cda9af829,5e904dbd73f67cadb131197389e5eea603a8f90d31e89d354571d853b5d36e5f,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,47.93814432989691,47.9,exact,canonical-source,profile_b7ffbaf8d89bbaac429d0b87,high,high,true,fixed_effort,,false,2026-07-24,fresh,100,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
xai/grok-4-5,Grok 4.5,xAI,closed,44,weirdml,WeirdML,coding,7b789d5e1cabd09cf83dfaad6aa79c1ef668a7ec0949bc40ccce5c0e5edd07d3,954f3123561471df21be35d750c0601e4654bd01109ff39a5b93c3d81b0ca7c7,2b35c9e6af9f4c3d2d0c9da17a6e48ecc93aa147e7c71b994af82185394750e5,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,46.43,46.4,exact,canonical-source,profile_bc00c246c21133ce08995d8c,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,epoch-eci,Epoch Capabilities Index,capability-index,553df55c267c79564fb7c6b416d015cadd1f09942782c7c06006ca1b5fb25d2f,75a8118f9165a00b185b6e2f66f85e15e8a3b1839b22d0e6f44033d6a99a654d,94af1f4d5b004df756ceffff6c0ae4ddf968ce83962e445af69cf84d736a3feb,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,154.22,154.22,exact,canonical-source,profile_438a0adfb228d31d4956122c,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,arena-text,Arena Text,preference,29b83070299a20d6bcfaaa8c79854fbfb04973d7b44994ca83af81195507e57a,c3b45739fe75b07836926a115a1c5f70e60d63c048120a15ec8dc287ec9b70c3,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1473.7034427077983,1474,exact,canonical-source,profile_4220a4a6f1311bc74f0c6484,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,livebench,LiveBench,general,a2c677fe20e9f166841972b53296a41f9cad912c58910caf3ae8832aaf69116e,15e63cb922421da8e22b5642e1fe2aeae3c2ed167ee866572da2610160bf6de8,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,73.13660714285713,73.14,exact,canonical-source,profile_8a98f0234f2d8eb2f775b7dc,reported,unknown,,unknown,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,simpleqa-verified,SimpleQA Verified,factuality,4d4425dfed18342519838c5cdbc0957d75b1063a59c4f16e483a8bad30be10b9,01384844761acdcd679fbfcf8e1e42fd6452b3e4df4b642ab033fda4020aa132,eb1583e703edeb823f406ba50e4126f2c85021e5a69a322f58b085a038b90fba,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,55.76730190571715,55.8,exact,canonical-source,profile_3a75cadcf4f028e0f62f9064,reported,unknown,,unknown,,false,2026-08-27T19:34:09.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,gpqa-diamond,GPQA Diamond,science,a7fc22dc293641d29df9369c1649f95a3d26312646c192a4de9d616d470188cd,9ce21ea579241cadce4279dd0798c5c886740d3cdc13497851b314bb6b7a9b13,1797eeaac0bb99111e90ece69a584f98d8bcdae220648f20fc59ff1ad76c5d97,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,90.152,90.15,exact,canonical-source,profile_c851e77e82debc794fff6ed8,reported,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,mmlu-pro,MMLU-Pro,knowledge,1b6b6d287a37ae899916b1e90a537d1c159db1f73877efa809bb93f9a58beba7,fb7091625697968065dd4e3bba4ff94f95d45a71681832474305861fce3573ab,0ef4fa4822bbdc7dd0c7790eae132108d2c2bc17b9843999c396f9f4559d879d,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,89.311,89.3,exact,canonical-source,profile_0e54016d3db95a8f54d1a2e5,reported,unknown,,unknown,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,be0bf1cce1d8e1e94e0703c4f390f60bb36211f9ccdcf67faad35afa102666e3,4f8ee2e5d94cfa918fb3a9ce5fe1edc918467d460d8b7a016034c3a86544387b,a5b341a2ae4d57ce93a4f520cbabc07ca59484bef7b4675d88563352687d0e12,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,95.55555555555556,95.6,exact,canonical-source,profile_b3e2c828e764c38eff97fcb8,reported,unknown,,unknown,,false,2026-08-07T00:39:02.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,frontiermath,FrontierMath T1–3,mathematics,5f70c4002615f5111a3959cd056ccf67a8b7b9f4467c858e78442b1965286ac2,61b59bd68901a84213ba31d199668291c65dc98a76f307363b77e61f7b739df0,664220cdf0cfc70a17f9aa1a388d66675817155056d69c7725a6617df98310bb,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,64.56140350877193,64.6,exact,canonical-source,profile_4a082360cdbfb26695a8efcc,reported,unknown,,unknown,,false,2026-06-13T03:08:22.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,frontiermath-tier-4,FrontierMath T4,mathematics,03733d880442dcf5b64eb5a7b6f9c082e47c8558f864897af6b62acfcff8cd95,4d0ff81e310b15d3cffad671549261918d521be552c0ec633a0f6e8a360cdeb2,792de66b59f8316ef0e0a3ccce21e0a763caa123578ca4cff217300a967b5b1a,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,34.146341463414636,34.1,exact,canonical-source,profile_898133c7fa458202db96273c,reported,unknown,,unknown,,false,2026-06-13T03:08:22.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,critpt,CritPt,science,f79dd9581706f064c216341d18d40e31ef38459045cb7ca60ed0c2c3f6a2b117,88551b7877cadcfc05839b5537f635216f76ca9f527b608a1bfb955d521a3c62,91f50faa87fd0d841f330a8c7b2b3a0bce11c7f101afbcf432c347910ddd411a,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,13.4285714285714,13.4,exact,canonical-source,profile_d772c2aef798e310edb6ef13,max,max,true,fixed_effort,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,simplebench,SimpleBench,reasoning,c49f7ed6777a9bfcf1a71e71dfd6b417189cd69291168dcc5074171e283c6c4c,7089813c757b4e0b019034086c36220ec6feb416b17f8de82342642491dc0a2d,29beccbaf009967b3fb5b2f9d46ebc0d6bb778cceb0260b854dd809068370284,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,70.4,70.4,exact,canonical-source,profile_73d6450b88bda658f569c477,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,chess-puzzles,Chess Puzzles,games,3fa090c062e9f1fc2e14356f5ad8788a38ff9caca6052783e95c417dc3d94ad5,08d38057c72b066dd3720affbc5a6595d13826b651c306b39580e852bda15acb,75149d26e5258c61b30592c7149b54f99213b246326aede07c4926cf81ac0ec6,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,26.0,26.0,exact,canonical-source,profile_a74bf66befbcb7f81dd6ab02,reported,unknown,,unknown,,false,2026-10-13T13:33:00.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,swe-bench-verified,SWE-bench Verified,coding-agent,b1cf2fa70dce2998079516a33f81d27be51369094dcfbb196cdd23f348f021b2,4e463f15a2658d06a97b6c6780a9f8964444941f63fad5ef4789583daa54e761,2fcb0eb9ad0659e6baebea2eb302f620041a4ec46e98b4ecdac7264d9ea2bdbb,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,68.8,68.8,exact,canonical-source,profile_8702741fcb32199baffce119,reported,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,deepswe,DeepSWE,coding-agent,1729584b15d82d058282a9ef97c57cf522f9d2c7b5d81e4685df92b5e5c0dc2d,0e2a61089a7bdf299d96d35ea04a1681b730f592a7667daf9ce0042a44fe6079,26f7dbbbd59c10993e81bb31ec607ed81e91dce338e64d5c3ca6702cda935db1,release_0f94a5c34a87744f357d4816,DeepSWE 1.1 · Qwen comparison run,qwen-vendor-comparison,resolved_rate,percent,true,21.6,21.6,other-supplemental,supplemental-source,profile_0d904d7fe29e536b30a70704,reported,unknown,,unknown,,false,2026-08-12,fresh,81,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-8-a95b-card,Qwen3.8-2.4T-A95B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,livecodebench,LiveCodeBench,coding,3ea1d4393a90e42b67ce8612404c0b27c0d5cb54becbabffb6d42658b1913359,9827af0c22243d7b1b3cc9cdf61e31c1ee0a0d29f1503a7e8b0592b08d5d4e57,337f5a3459785a096ef74e3c817e6bd9f080dd141ad6742996950ebad24555e8,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,87.057,87.1,exact,canonical-source,profile_e9712f59624c456abd520aa2,reported,unknown,,unknown,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,scicode,SciCode,coding,38875a5f0454de995f47ef262cbaf54107f698a3cf6567a7a4618bc048b1d207,9fbc0fe5397de6423e3cf02ba3c75d707436ae67dbe05081ed771e8ff84b076f,c9f86708fcaa36df1d06c5b20b1487ea19b6b4ed0166ce1d63b6cd30bb0ca964,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,49.537037037037,49.5,exact,canonical-source,profile_d772c2aef798e310edb6ef13,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,ale-bench,ALE-Bench,coding,ac075713df149595a16a9cc24cb902d6724d9e437dcdf2f4db476323ab7bfbf7,ea0007d8fd8489216e0e73f76a5147aae00e2a86a90fc4d1f9f185e3dbbf5d79,921e8716214d093a8a220fc7e1c823ee9f3af73c6aebebe12122685d9c21764a,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1189.4,1189.40,exact,canonical-source,profile_d772c2aef798e310edb6ef13,max,max,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,vibe-code-bench,Vibe Code Bench,coding-agent,f5ee8ab7fb7156f8a8fec8c3fab3742aa5bae0f5fde36fa4fb3aaa90209af701,b0dc329898e71c57a427f461919f2c998c705262af7a71564d058c1582fedc9b,e45296bdd0fd9c48c3a514a81fad4d556c572fdb2f97fb2c3505e03d334e5532,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,47.671,47.7,exact,canonical-source,profile_5db0d7d1f8becd7348837d1d,reported,unknown,,unknown,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,webdev-arena,Text Arena · Coding,coding,821cdc5a846950737549e1121edaa7a732f27ad3b25851b60cd5c1f1a3024af0,a054adb885a9e3112136eab9d72fc169da9cfe1974803aa31bc262dfb530d29f,c221be49c9f7b24eda837bb55fb57138648e6cb4ad2bc348a4752587649f742e,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1515.1361678106236,1515,exact,canonical-source,profile_93f8ad037c4bd04e2de7d0e0,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,09535bcd6be89dffe9eb1ffcb9e47746beacc76e9f85dc09fccf3ed70b82598d,1db710d74c3f09ca5c08868ebce3bb79a7f6a5e0942b5a57d2b59a6215eb5200,612db9dc822ad0127ca02d953d2f00027f18f4556fa9f2ebd8563da305954e88,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,61.049,61.0,exact,canonical-source,profile_9834bcc9b3e87e87a66577e8,reported,unknown,,unknown,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,skillsbench,SkillsBench v1.1,tool-use,0b4458942c5517b7ff7da7a12ca1e61d96bd31558641f769e07e419205119317,e365ccca63745a886ccffc98dd77d5d3a48460c04323f088738897867f70d2f9,5a87a249627548ba41ab1e22adf8f465931b323f1bc4cc1e11b5f7988d5b1a3c,release_ff4a6e288c3c17d588478b46,SkillsBench · Qwen comparison run,qwen-vendor-comparison,pass_rate,percent,true,61.2,61.2,other-supplemental,supplemental-source,profile_0d904d7fe29e536b30a70704,reported,unknown,,unknown,,false,2026-08-12,fresh,81,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-8-a95b-card,Qwen3.8-2.4T-A95B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,arena-agent,LMArena Agent Arena,multi-turn-agent,bc8ff902d0ff5f06839ea27996704cc3d2d8c536e945630954829f7c707fbd7c,82961265780ce4c7bf0c2f50f19258859fdd639fdd19477f0d3a731f0df661ae,15f5fce34de6357feca787e195d1553ff7d07b9ff955fcd842be20dd4ae903a5,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.06586293614361291,'-0.0659,exact,source-specific,profile_a11f0381d740fab28f4a990c,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-7-max,Qwen 3.7 Max,Alibaba,closed,45,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,74126c7317d5ac2412c940c3683e864c121115c29308b2aec54693eab03542c5,6861a84590257cd083461d7016e285e1f7e6510c904d51b571d6cfbf537ed8e0,6472cc5c76109c0e8c0e0bf720a09f008491e120de73a07384de2bfa0c1b6aea,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1135.46,1135,exact,canonical-source,profile_b653e6b09c5d08a78703143b,reported,unknown,,unknown,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,epoch-eci,Epoch Capabilities Index,capability-index,e6c673d0e64bfc83fc0ac33aeecc04c06a9230a6b70d09f0be4ffca197559a30,7d3311469c824d645215c2b592a0de27e3626cc2989e4a61e5e88437742b2c8b,a2ccc4b27dbc724eda9bdb5a62a3d16d84fb3c9718347a54f9a0639ada6f9158,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,154.18,154.18,exact,canonical-source,profile_cd729d89537d94a3078f527f,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,arena-text,Arena Text,preference,d1070ce488746380f495b6ff70bebfa47ec29007a3969959dc9fe19e45f295f0,37f4bf97dd1c683b5cc08d409d96fe7bad341abe39465c9a8153c04e40c37a4c,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1490.2351603730194,1490,exact,canonical-source,profile_99814c21bc9361f199e3a520,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,livebench,LiveBench,general,fb772e95f0047b71ea0dd7da8191268ee220385e52384101b827f8295acfe03a,ffe5b94b254a3e3ad19d584830b2396f68ae50071888153db32f0bfbc458935b,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,75.29954761904763,75.30,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,34272963f1a255e23d3a7525f678a28edd208a2a37fe40e21377f77381fb2c0d,4ce906c27a5c7e17f069b59dcda1c7b1cac75562cecb33ba752bd7ecc7fd19f5,4af2544e468228c0ae8c94b37bb85399d9527d09f75350cc2badda8609a813dc,release_5f3023448c2ae4632fe1a43b,reported-runs,public-multimodal-mixed-run-configs,accuracy,percent,true,52.2,52.2,other-supplemental,supplemental-source,profile_1cb2c5290629fc013609d211,reported,unknown,,unknown,,false,2026-07-09,fresh,115,2,B,2,B,1,A,evals-report-verified,0,evals-report-verified,evals-report-hle,evals.report · Humanity's Last Exam,https://evals.report/benchmarks/humanitys-last-exam?tab=scores
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,simpleqa-verified,SimpleQA Verified,factuality,c34319d8c898cbb5e90f8dc6c8615afa231b02eec3807fca8c0d1ad2573b0a88,28c390ef12dde21b3a10e1d59c86e6635087105c9d3ab1f9a58ac9c3efd8dc1e,c8e8fa50c510fadf6b63e74fddf3368f09db7944e2accd9801f2362c09998e52,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,57.78894472361809,57.8,exact,canonical-source,profile_a5dc1bcf8ba60512cb3ee6a5,reported,unknown,,unknown,,false,2026-08-31T19:21:58.000Z,fresh,61,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,gpqa-diamond,GPQA Diamond,science,aa68bdeda3cd6061177600250fc1c7d5172c68abac197327f254c959d5ad8bc4,ed0c328a1ecd7a93c09ebf6abc135b0f08ae5b672df5266cd1ef6926728fb6d1,068ffd607359f63291ae31135ffbe94736591d10d6cc4f69a3b1faa658f51820,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,91.162,91.16,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,mmlu-pro,MMLU-Pro,knowledge,6c2f8b148ea6f56e86f214d24e80af8db2be3f8714bc8c46b9fca70a3533f4b9,f2a4a2e38323ced376bad1846546955b9fe0a31b548613593b024ed67b0f2ae4,fb8ec0afdea20c28468a51bdde7f83376d0cbede8ce7eb3509c5d99913278743,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,88.726,88.7,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,mmmu-pro,MMMU-Pro,multimodal,31e79eb48f751dfd9dbd0b634dfb35765fd1608db2c7f13a80be899af6017bce,51fd66eed857a3370462d059147291b667a316334177d4fe813e9cc04b16bf5e,50ab376774c80f746c4523aa636d752de12d7fc6b69dd3f69d8c98eae6f338a0,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,86.59,86.6,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,critpt,CritPt,science,65ccc23508a058cd33abe9442a7a511b3e5d51190f830a8ada599541f9c158b9,d226c947164199cf42b44af140c8bc9d09d281fde7c0d07c1b352010bc219edc,2aa326f2519e1a10af2940c4581d65a5efdd1f244f87bbfaadb7feeff638287c,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,15.1428571428571,15.1,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-10-08T22:23:13Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,chess-puzzles,Chess Puzzles,games,fcea5cf85b36c2dd99990bcd892111a606c8fd40d319168a304bdc5978e37579,4b35b204fa8d869012f980f716a020bd32436a58b2722db81feb8e5354ab7ef0,aab014abddad39d3755c5ecfb367cbcc57cb4069af5df8d584eae557b8f92d69,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,34.0,34.0,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-10-13T13:13:33.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,swe-bench-verified,SWE-bench Verified,coding-agent,1c62bc9676a9ff47b8cef6fb9d688b169119b8b0d6d108e93a67e430e6ddbf71,1767cf26afda607d2032caf009207c778e5de78681fafdd37a9d8ea3dfef6860,17377d55f847c7d6a67b1ee45439a926553ef076ef7167776a5f15c416253783,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,82.0,82.0,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,deepswe,DeepSWE,coding-agent,fdcc066f6b139ca82842e9053f1ff4383019e648a3c2a588c39aac61d680c8b7,d1c241e03c555e41c0fa68926303824db73297f307756a1d8a0576db8f0dd4ee,9fa9b172911f42872d2217a905d99d07616d897ce35fabb291c5e9ab2403da64,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,53.31858407079646,53.3,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,livecodebench,LiveCodeBench,coding,deab316c81419f3aea0c113013cedd5466abe56604996ba01d9b4b631ec17a19,1f312b912ed268ef06d2a90519e2d1c8a3dd9d8e28f44e16a99f2583303194a7,59a5b7be6da495b040103b021a1254a911407f2c886a2d5b0c3ef34c196647e4,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,85.906,85.9,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,scicode,SciCode,coding,c1dac08f7aa8a08b63fcae6753dd3f31ef3402b8465039f67a9f6299636a1734,060dfe1e3046a4def1a414a1bab1deffd10b987c6710fa90963e8c744778da82,561c3b5e0b6be568f854f8f88928ec242f42067d7899ef94ac31a6fabb18b057,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,58.796296296296305,58.8,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,vibe-code-bench,Vibe Code Bench,coding-agent,8a0bf6b5046ce535a59a7770f2878615aac020fd8b2a5a37acaa3e687637e0aa,1d936289bdf0ed5b4936aa301e7c5362b1522d30e488a3909f63057d03e46c14,e6ce7565d73a1482a676ee73e6312177730d24a3ce1b883d795732ef36940a29,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,72.158,72.2,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,webdev-arena,Text Arena · Coding,coding,67c2a14ee0be74fbe6f8067cbed3f025abe3a8ef1346cf088bb0c092469be244,ccde926dbe3a6c162989f74079c0eb5895099f2ab44968fe18a581de274e9c4c,40fadaaf9f417c1c4b0741af70d570fb003240bb0230c2a10878164fafb18dc4,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1542.4822191942499,1542,exact,canonical-source,profile_f91f386d6cb697210b7a2893,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,programbench,ProgramBench,coding-agent,446fe023e1b7ac411f5aab7f85971387388ffbd27f3ea6317d81f760d1d2e459,5aaa135d4f5c698462e9de7a6eb809986bfb1cf17560f0a276ea6ffcc9a2730b,b7a5c952d8d739e3849fbe90c2d7bc13c10eb2bc7db560c341f0bdc99bf2cdac,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,c65f3da6a4cadd2a5177b4c94dca1032ed1914c34d26004ab7788748b2ab1956,a9ea18b5d38079cb47ab4ad83151abedd59d478c99281b39af5857827ab4884f,e920d28f173952295012036e458864bbc3e92afa535baaeb5a6049ff2b089e00,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,69.288,69.3,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,mcp-atlas,MCP-Atlas · Scale,tool-use,186cd5293c01acb4f6656648f41733d55f5be3dc50e48ffc9992d6ea39b5c9f6,35e14bd5a3a2095a4daf840fa12cf2d8008148a91d35f1197912552ddec836ff,bf94b57049d0131c71f8140b80645e0ff207cd72060929ddc500220c48f714f9,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,88.1,88.1,exact,canonical-source,profile_f46411f67d3ab003209ce38a,reported,unknown,,unknown,,false,2026-07-09T19:46:54Z,fresh,114,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,apex-agents,APEX-Agents,tool-use,8df08977f055f2c498938535fd37f20ec655e5d0ee9eebbded6fd44921b91a3d,ebeee9dfe22cc30fd5f983f865c6edc89e2832d8f11a171266abb25f8de46a1e,f1f64516a91516ef2dee6544987d37bfe9ce300153eb3effddd38f983da5bb94,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,31.8,31.8,exact,canonical-source,profile_458ad66cb249e06fdb6aaf95,reported,unknown,,unknown,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,arena-agent,LMArena Agent Arena,multi-turn-agent,027c38508ee1ee2c04499adb29941a32e9be5f834e99b9e3952cd3969e076373,ae5019e3e316fd2d0f75ca745966100e411b9b3945ac7d1298eaa970d8f02afc,60b01f02ed9c4c36405f68356104ba7cc7a9781f3af21c7794c7a518e5dac833,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.0766934579263557,'-0.0767,exact,source-specific,profile_60c2da8ae56f193d786d0708,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,vending-bench-2,Vending-Bench 2,multi-turn-agent,44d6176cd2ba3f099a6d4eaed90c4e94db0f61c15e636a808345a20a35fb6a3b,354a74bc97c6af902000c7cd684229ac992140c89f35e0798cd1ac16a85b83f5,b622d2048bbaf7ec329beaf3505277f8648ca0c8997f8289c3e57a872ac18f59,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,6520.474999999999,6520.47,exact,canonical-source,profile_0cc4cc6b4e71e0063d4a49e2,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,2fee9caa64142e299728bcb1a9a29a0afe1ba17648f18be44ed8b117b5e5910f,c2e24c0eb62e13612ef551a9a0128bf3de14496d0c1ac83f73aa56dddc6aa7ba,451b63da8bc955236d2c7eb34ff19569bdbd119e0e1bea553a226cada2372c01,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1217.25,1217,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
meta/muse-spark-1-1,Muse Spark 1.1,Meta,closed,46,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,18bfe3b62cd0d1f61b29dec7f5683da1cab38cc79c56d6e10a86cc18df622d66,edc6b8eebf82c09582b9f9c9d00cef270186d741b7b42ecbbe81e24180e6dfb7,586a45dfded2cfa3639c1530e9bbc6ff3221b066b8305b9464d7b46e95b10f2c,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,40.4639175257732,40.5,exact,canonical-source,profile_f65a7f664c0eb4433d5d5397,xhigh,xhigh,true,fixed_effort,,false,2026-07-23,fresh,101,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,epoch-eci,Epoch Capabilities Index,capability-index,d88fe00aa9881f948e91d3962891a54b3e3bf3dc75d7bc54e0ef13c550664667,902d041af46610473b3fe7c5153f3495b92cd4feaa0b5530f5ee92055ae2359b,ea5f75f9769e9199a30600f9008b8698432bc712e923c5597dd5e4a8d3b46d86,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,153.83,153.83,exact,canonical-source,profile_d195c64b8a3ecc15e17b76c2,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,arena-text,Arena Text,preference,cb2a9631f67a0b565544419f4c120721cb02dcb13845f842cc54973e719859d1,4aba70467063609624e6d0177f70d3227f9d5ad07ff186a766c3efa4f6413af8,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1437.2916252001148,1437,exact,canonical-source,profile_da6fffb3400932260676d0d4,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,livebench,LiveBench,general,e3caeed6cb3d9a4667b245a572be8879e4c28f5c32655439862ebbcfa0be4ae1,31b49cd4e2feb2a2849ee3fa16bdf57c2ba8487ad63c74ce83733f46b52c96ee,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,74.6349642857143,74.63,exact,canonical-source,profile_da6fffb3400932260676d0d4,high,high,true,fixed_effort,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,da7a887c488e329e6285baac58d4eef1981eb0b96e7f28a3533515c7ac5a4757,edad0d240b65afe29586be2e182b2ac61912fd7b46b14dd2a63eb203b55f070e,8fcb2b3e6061c56477be2203a02c1fb26599a3be145c6bb5792519f004054328,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,27.8,27.80,exact,canonical-source,profile_86dae51067db1dc1fc5309bb,reported,unknown,,unknown,,false,2025-12-15T23:38:26Z,stale,320,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,simpleqa-verified,SimpleQA Verified,factuality,0886da10d91ac72c56e07ef5956734fd2a59d68512a4b2add7fff885aa613bef,d39ca95785e72a0d40722b18af019a3050468f3e76a4523bcd70a9ec74922be2,222ed766bfebee500cdc2d8668f454b6ce0017d1a7ab2bb2f239ab09339d5374,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,37.1,37.1,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T19:27:55.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,gpqa-diamond,GPQA Diamond,science,87f7e8712530d46c9281197219988f63ab13829419f9907786a5dc489809e91b,8ee64d7a58e7bb289c9d96f013a3f1f723a2d68c11a80ae23cde89b801870187,4587329be7bbb42eae096650bdbe2d015876fd754bd8beef059af136cf9c3438,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,91.666,91.67,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,mmlu-pro,MMLU-Pro,knowledge,f0d3873f45944419590fd7bab6c806ed556989fa9bc61702da0e28145f97ce07,09852e0c41526b5fff3342e1e110e1dd68f718a2589c8892d425ddb8e99bea4f,604c4461ee4f7b1a182791cde7471d23b2cd64d0acb40c36898dba53b6a141ac,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.234,86.2,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,ad6572c5e09577fa86f5eb101835e6c8103bdc6b1ffebe3d93e2b7cf48d385ce,46c91470889dbf53395c1f0180ea2a0bfe809ee3bfd65e33ced88e3e298dab7e,262d18a12360b932973b89b9d03963035c3c57ce099ed5777db7c6d87332a7f1,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,96.1,96.1,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2025-12-13T15:56:00.000Z,stale,322,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,frontiermath,FrontierMath T1–3,mathematics,7f1deb646f95d7c82c77c9d26b2a5f0cd938c908449b20821005af5fc1fb2f87,746694c238eab19c468b29d852ac92bee0b678a81efae5b5cef89da836790c5a,5686c6437aca0367bc365c1104f9271197e9b9dab735a9cea4972f021a45bc40,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,67.4,67.4,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-06-11T02:48:30.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,arc-agi-2,ARC-AGI-2,abstraction,c9ffc1648edefe35fb1bd6542c4467b15b8a9aa6f547b76dcc2dee40dc48ee32,fb35af866b01bbaee5a196f91a7d9eeef62bab5621accf41bd29382a9f3fb920,d01d54cf6663c2478759bd86f00cfc3e917f8de7551571b190f30e5bbafc1a62,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,52.910000000000004,53,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,frontiermath-tier-4,FrontierMath T4,mathematics,5fe887e4fe0776f8611fd462abbd260b63a382108595552d22db284112a1bd0c,61d7a1af699c9feafe0df5334b96c01a4d06c944a3b2b640ab1d4a5b88b3669a,549bfdf41fd03bd148e9ca762423d3d6b133ad71b87a65916b717b0b26fa4850,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,31.7,31.7,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-06-11T14:48:04.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,mmmu-pro,MMMU-Pro,multimodal,184099246db7e2370cca873ba5962eadb7efa3271fae0299d790f20a39cc8704,c2e074776eb6a0738d2e174e5d6959bd4429ea4b674542f9bf0011202b10e821,a33dd7dc139050239d5ef09ebfbcbdf12eefb661c41d6a4da238a318313fb8e7,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,86.667,86.7,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,simplebench,SimpleBench,reasoning,cf1430603d6ba58d862e3622fcf8fc50cf9881c28e6e7e247d564572f8a52dc0,0ac7e166f6adc30adda7cfeb3db134273aa9d707ae2d11fe1e01d12b21e5a5e0,3f440b29a11b0bbf50a91dff1d673b2869ac88c6ad8dd8682cb0a9fcfe3be965,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,45.800000000000004,45.8,exact,canonical-source,profile_da6fffb3400932260676d0d4,high,high,true,fixed_effort,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,chess-puzzles,Chess Puzzles,games,e0a04ac86a75d678999e7fca4acdcdbd899a11fa606e2921cc7cf79cc8da779d,574fa137654a157be27f9f3ffd50eef211fe7321633c8b06fabb1aac6cc99d2f,81b8d14850973a8d95f6442d108a5b2c736512f2bb68e3d0fb5f0f162afd9a98,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,48.0,48.0,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T16:16:30.000Z,fresh,17,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,arc-agi-1,ARC-AGI-1,abstraction,2f339099068bb5dbe440746c19578cf6cc982e0eb1a6094576ba43f8a09a3d47,618dc545255240bbbfd490c6c655fc0fc3e983472960f07f96456364ea59cd4e,5093acc4b109689bb8f30a930c66e72ace881fd78b814418150bdf331bc1ef81,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,86.2,86.2,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,swe-bench-verified,SWE-bench Verified,coding-agent,2594b06434c2a2829b1b63380f758557500414c4e6966e9b7416ad3db5966070,7280c518a6a3cde615a0a368db9691ac5e383b3da967af99fb9a0b90de1c0316,5b48f541e6e08a71551f7ee0691e079128f29a5c08a630fe9b7f6667d3d53219,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,75.8,75.8,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,swe-rebench-v2,SWE-rebench v2,coding-agent,d15355f780b41cb618fbf22675a85d380bb951e86e22e13702f1430428ddf0d5,2d2a0b7e33c974264cd85095cd877b3081a195aaa5b93699a2a941cd99d62dff,f288398b66ed709eb0d9a928ab5ecc2b376564d2d163f6ecd2518791219528c7,release_b5f5937b2cb1e2b756288948,rolling-v2,2025-12-15/2026-03-01,resolved_rate,percent,true,59.01526992508781,59.02,other-supplemental,supplemental-prior-window,profile_9c94798e73be04b03372ecd3,medium,medium,true,fixed_effort,,false,2026-03-01T00:00:00Z,stale,245,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,gso,GSO,coding-agent,e1adffbcf0347ea927f8e3e7cbb8f559b2c328f4734dbabb8f6f49a44bae0afb,4ed1dd18dbe03141c9e92970a5e69b360fd8661b3ba8c77d148446e0c36fc6e3,5e521ae728dfdea359144507dde4e6beeafe8b17a3a8ca8e8abbdc88798d0ee9,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,27.450000000000003,27.5,exact,canonical-source,profile_da6fffb3400932260676d0d4,high,high,true,fixed_effort,,false,2025-12-18,stale,318,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,livecodebench,LiveCodeBench,coding,32457c5f40a2ba8208b375173d1916ea3e9b0930e3975350a51d1e21b0e4f14b,b40e0212061b0befcabfaee0cb39cdaeae6296ce9c84b7b9fd6dc8655eb1db7a,83361460056ca95d46080cad51bd4f4b367db0cef05d53c3e86940c46e8c2075,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,85.361,85.4,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,ale-bench,ALE-Bench,coding,dc87caa3c7ef5ad2a65848c220d3102c7af01e21495ce3ae83dc6e72120a7c35,8eae3ea9c9b1359dd15c0b1315acae669e2f330b9dd8b41f8ab5b9cb8c4f445b,f0a4cfc02667cf63bd3491e1b679861c8ac630ab58bfea090534396c2f2724dd,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1293.55,1293.55,exact,canonical-source,profile_da6fffb3400932260676d0d4,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,vibe-code-bench,Vibe Code Bench,coding-agent,d065565414917453500ae80a2ffdb6bda021c822c7c693040cdf2ea5e5b89307,cf6c8e9748239a619ffd046c086ca21c1e71b83f63aee8687a8895af49cdf607,27296a73811abfc9e4144c985fb38423385f57fc85d2fda19967dcd197ace2ac,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,53.499,53.5,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,webdev-arena,Text Arena · Coding,coding,67e45f436795f57f90bb8297b8e008b81f35c4f6286ddc12fa755cc9e63cc1a9,6292db1cdce1c2d6468f0712ec0cc72b448b81a06d7be82328ef34a4fb87e5fa,9d5228690ec00c526f3f023e0ef249f362499bd555fb83c92c1d3658117c87ed,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1416.424922259506,1416,exact,canonical-source,profile_6393519d4489cc2ad96abfdc,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,weirdml-v3,WeirdML v3,coding-agent,037883f54c632adc081fa512f50902ac937ba67a324cb3b0519dd794c359e512,79d45ce31ce4dc700a13b022458b05df29045422394c7b87325a97564f581f8a,19e0b2da70b62d41229da1ff68c86ab5dfb7ed43b789ee18dc77b8e1faeb521d,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,5.00787722943723,5.0,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,skillsbench,SkillsBench v1.1,tool-use,4426c94fd8fb35d6d48b96094b86ab3ad420fc746aa7465c1ac352235c0431b9,e8f300a63c2e28ffb5ad2d802aec38f172b7a87b821113d7da1cdbf026f51851,de4ee8be7ebbb4075f4b810c6a6dcd2e1f701030bba81db0d7f2320352479107,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,51.7046,51.7,exact,canonical-source,profile_5abeb85214411e952b3bbd17,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,bfcl-v4,Berkeley Function Calling Leaderboard V4,tool-use,cd11e4923ff70e1c795e4bbd90bfac5836b96bc3614a13bad7820f0e2008e9af,4be67636ef17c1b568b6dfa8779ab07fb926850d78ce4585cab0018370b48d3e,6a9c99bccbca11748bdf70a2bdd6eb864726c94bf4a029f84ab95898fe96d192,release_7d2560723dba9d549cd10b25,v4-ede5081a24bc,overall,overall_accuracy,percent,true,55.87,55.87,exact,canonical-source,profile_c45390ebff98a31c8653458d,reported,unknown,,unknown,,false,2026-08-28T22:03:18Z,fresh,64,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,bfcl-v4,BFCL V4,https://gorilla.cs.berkeley.edu/leaderboard.html
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,vending-bench-2,Vending-Bench 2,multi-turn-agent,60fe15ef9c130cf445f0c84d1a117f88b90cd5e373acfc0be0abca9d92bd3df6,1991f985c7618b1227a484879f8260b3d070aaad5aae3ffc4e3e321c404e5d44,6da35d09238c9df45d2f323fe5aeb267713f7fa3fa3027c58f086ef6272f3446,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,3591.3259999999996,3591.33,exact,canonical-source,profile_71d5b9dc73fff7d1c5cd8972,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,gdpval,GDPval,professional-work,ef58c0e84c6163670f11b4d8afe1caf95b1d452a09b57de0b3c56ff5fd8ab42c,dfc3c202922dd1e2ff81437d478664ab26175319fdd54e929f38338d1cda4a13,5a3fcb25a7d145e0ee5aac9c23e0f8e4c772626f45806ef93461b4abdb803807,release_730527fa346470703bd19e21,GDPval official leaderboard,expert-comparison-win-rate,win_rate,percent,true,49.7,49.7,exact,canonical-source,profile_ff3aeab8b13f468a67900233,off,off,false,none,,false,2026-09-02T09:27:05Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-gdpval,Epoch AI mirror · GDPval,https://epoch.ai/benchmarks/gdpval
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,15c9fe172460154f3d0f3483b7f317188f1ec54ab46393d46b3c6f3448b954fc,beec4780e04627cab69a6e0396c1852621dad5f4b314dd3d126aa47863287809,fddc1da1ea5e3ec0812c175281f75e6b59702134f5a28ebade6fbbc00b5ce9fe,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,32.22,32.2,exact,canonical-source,profile_da6fffb3400932260676d0d4,high,high,true,fixed_effort,,false,2026-05-05,fresh,180,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
openai/gpt-5-2,GPT 5.2,OpenAI,closed,47,weirdml,WeirdML,coding,21203e7dec2f79cc953e15809672b81edefd4ae82dea9aba0b70a93348278134,728b4ba4bb8e86cd9e0b02f17c57dfbb7ed5d0b5e6a1e1d422b39e5613ae5654,f0633360f72acc77df9d01cf9d2953ff8c176ba7ef401ce2b267e43c9b253c92,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,72.19,72.2,exact,canonical-source,profile_4a86c49777c47cb202b9dfc5,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,epoch-eci,Epoch Capabilities Index,capability-index,bf24a9a33e5c8e5afb92eaf3ec33b808e00fa37e54a64e474dbd4ec377a87d9c,c82d3ed348bb9eca185213042f6e641fd0a196e949e56ee0abd86a554210ee69,696f01d1b678fcf06fc846be70cdd27cfb20b9cb6a5c2b9c8158891a38622bf0,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,153.22,153.22,exact,canonical-source,profile_21273f06524672cf31a6a06e,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,arena-text,Arena Text,preference,99e700a4d333d22581147d91de6eda84cb092fe3ba15ed8ecf952e7dcb21e37e,a94aa82d0e514ae92b4faef149e30eec21c8b45ec12cb6755ddfb09b3fc31c8b,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1485.4329182845129,1485,exact,canonical-source,profile_0271df1781497779ba7d4b50,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,8844c39345c641b007ad9fd918b5dcd002421d6b51d77815b44ac91235a65a97,4abb913611c36a2d5d506f3487440581af8f569db2b44cfabbd3dd0ca0f1c5aa,00ca9c954eda6fb0df85537b8300d0bb776f4eec980a8b7cf500193aa43796fb,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,37.52,37.52,exact,canonical-source,profile_b228b8ba7c4906d4189d84aa,reported,unknown,,unknown,,false,2025-11-19T23:50:49Z,stale,346,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,simpleqa-verified,SimpleQA Verified,factuality,131f109e88713c4eeed3b24b4f53909c70536e723213e56005e53576617a284a,3753b61f36ec3a84e7c5fc4e68b9ff31069afe44c9bb02811545d2e1ba2d9fe3,b3f27ff774166a33798e6609b791b03ff7b9c2a1ba3229d991c2865b8aa97513,release_ccacd6f679c9142b534364b4,task-v7,verified,accuracy,percent,true,72.10300429,72.1,other-supplemental,supplemental-source,profile_99f15e2ba246f14793e1e251,reported,unknown,,unknown,,false,2025-11-13T00:00:00Z,stale,353,1,A,1,A,0,A+,official-benchmark-platform,0,official-benchmark-platform,simpleqa-verified,SimpleQA Verified,https://www.kaggle.com/api/v1/benchmarks/deepmind/simpleqa-verified/leaderboard
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,gpqa-diamond,GPQA Diamond,science,933b8f00c922085ffffc476de9604a8ee75f9fedd83fcfddbf5f58ce01eb937d,1757d5b37dba58bc958ea67a6d67c76c0de56b3a2f99dc4212b5858ba3c55a88,0c969d64a5adce4cbf73e75fed860e77cb5d6a0c296fe2c63652882e8ea2027d,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,91.666,91.67,exact,canonical-source,profile_adb9d064fb6ec60affbc6752,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,mmlu-pro,MMLU-Pro,knowledge,b1e3ae907e658376bd5439a5493ecf1e438e6353eee755321be614faa32d6d83,429612b980e9a03188db52e44b6bd4c2c31bc5e9d68a45bcc4ca40f752972816,2906ce753f69028b9c07a9f314a3c07a2a1d0b2bfa51b5d2dc7b948185013c3b,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,90.102,90.1,exact,canonical-source,profile_7c54b62064912898a8a23a9a,reported,unknown,,unknown,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,ceed1a5499b83d764aa861848a89af69bc4ff3e8b79b265fcb32b96708d04134,e494addb2004393b887639ea4910bfd29372b67835b4d37202b5df9cc979facd,9dbd0a5981231b06590c4da0465a446a5729320593618e11466b8ee4c9e19f03,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,91.38888888888889,91.4,exact,canonical-source,profile_2f83be29a9f1348d44b039f8,reported,unknown,,unknown,,false,2025-11-19T15:16:13.231Z,stale,346,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,arc-agi-2,ARC-AGI-2,abstraction,c6f0a222d2d82a268408e5dc58daadcbbf7f6765b25869f11a401b25c3e91335,839e4936351061d3ef52eb7d00593f4348a7670a352c142fa8c8f97efd560aa1,99a1d91d545a77cd7589ab38d5e6c23ee720a21c66735bb8e5f3e86f432a91de,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,31.11,31,exact,canonical-source,profile_55859642c21787720a9cc539,reported,unknown,,unknown,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,mmmu-pro,MMMU-Pro,multimodal,50983d69c7e0dd7cdb65ba3cd42ad39702da381644f9c633323d49d94095cddc,14bc5cbf255a4fcf023de6f5acd2657ff8f213c88fedbd3862fce58cdf7051ef,75ff959ffe948428097e0bc8e42317d066196c8703d86c6002167b22f98e1184,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,87.514,87.5,exact,canonical-source,profile_adb9d064fb6ec60affbc6752,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,critpt,CritPt,science,7259948453499b4563bb3c09f41849ff861ebfce6ae6dea4888d1a90100b20f2,5cef0609964e61f49a2286f2309e4813e869226bd44e93303ab1f5206eac8926,9162a8226815c167a678857e1353c87c3a641643f4952f0fa1f5b1558dcb125f,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,6.9,6.9,exact,canonical-source,profile_06cf7a8f0e7a9f715887e058,reported,unknown,,unknown,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,simplebench,SimpleBench,reasoning,78fb9d1271b2d874a29e85f54dc75a06c061ac92650cda0914a4cd0c802457ad,fbe8233ec0de78536a6be1d991750038b3b3379f51d3b676033d5042fafb8893,5598f250eee0f48592d2053da9bdea88c7a82ed2e827eebbbc2bc6954db727a2,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,76.4,76.4,exact,canonical-source,profile_2f936e9456cd58720653f4cf,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,arc-agi-1,ARC-AGI-1,abstraction,f0bfd46c1967002ae32fe5e09f14637ba7fb1f50fad5b4d94b16bb86faed986b,3ad436a4ff128e487d40977f78ff2872b070796485c20ad115aa4babf83c29fd,a61cc4282fc3788ccaa14ffbf2fe051a812472f438d814c2cfb9795c40923e80,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,75.0,75.0,exact,canonical-source,profile_eea65165256ff8318729c400,reported,unknown,,unknown,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,swe-bench-verified,SWE-bench Verified,coding-agent,f9dc7da3677d2191ebe5bfa1505841ab78f8399e31afc8adc269ae2cf4eb08ee,f3694aca98844bbcd7ad132ff053e46233cebe650b2073a1bac79bf5f104590c,f328eafabac495599c18cfa83048fdb84b506a7b15f398ca693d6429ece72384,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,76.4,76.4,exact,canonical-source,profile_adb9d064fb6ec60affbc6752,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,swe-rebench-v2,SWE-rebench v2,coding-agent,08ad310b01ca4ef1f0154b5b5c62604db450fd74b5d33b118cb07bbcddfe6fd6,f4e92d07c2e51715b7e2c507a4278775841d0f1a1277c63bbe9e1e32eb28c8ab,aabe485e331299acc0f7029513ee3a87f1b24a21da19427abfac925db891570b,release_1c18f4398e32b2e0fe7f7d86,rolling-v2,2025-12-01/2026-02-01,resolved_rate,percent,true,52.74342105263159,52.74,other-supplemental,supplemental-prior-window,profile_2123e902a08dee70792ebf5f,reported,unknown,,unknown,,false,2026-02-01T00:00:00Z,stale,273,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,gso,GSO,coding-agent,a815a2e0b125c7f2a75fdf196f1a7dbbcf0e0e29e557d514a67f02df4271fdd0,c3af8c516bd038a63718a2d1980e5726f3c63c9c5980481cd2ed0b334ec581ff,93d2c24bbb66465156fda6199944da881b585d0d0009827798ac52c8d19539c6,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,18.63,18.6,exact,canonical-source,profile_f7289fba54bab2a09b2c9891,reported,unknown,,unknown,,false,2025-11-28,stale,338,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,livecodebench,LiveCodeBench,coding,b9bea7aee720be6a1e80bb1ee84ba6d0339f749f54e51ccebfb133dbd8e7d992,bea9ef226c9e087280ea6c048f12db1425e17ab812eca07774b802b54214d33a,bc9d2db7e409a56aeee431f3e4d3630507bfcfdd8e727f129319995a6df577fe,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,86.407,86.4,exact,canonical-source,profile_fef4a508fae5caf5110e0767,reported,unknown,,unknown,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,ale-bench,ALE-Bench,coding,1a37b52dae7448b9f7168db52fda20b906fd7e007097635c6a2136b5935ac0de,46e9b31947bd643846518ad2b345dc81c0345df3ebea2cbcac50e647e0eb4a6f,b90b91119d6e1c0d23d68ebf852ba81fd02dcd362d1175a124eb4f432f65cf7b,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1176.75,1176.75,exact,canonical-source,profile_adb9d064fb6ec60affbc6752,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,vibe-code-bench,Vibe Code Bench,coding-agent,a02cac52083e378a76574520a2892555c8216d14d3a5c574f1d1a5ee78256d30,ee8f71157d45511f4f610a86376e77ca6c964b3252590645487dce53eb2aa379,6a1a55d7b3845f0e6a5a7f34dfae97f692b4aaf56df868dbb014fdd21a7cea5e,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,14.3,14.3,exact,canonical-source,profile_adb9d064fb6ec60affbc6752,high,high,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,webdev-arena,Text Arena · Coding,coding,31e257c47e363a356f8cf64c7cb2cf3cfe90ba5c7d08622c378439c1fdb00de9,0d8c1096dee778bb7e42c88202be3177ae2507b46cec9102500dc790f420733c,df2c2f21fe2bd96c34853b96b34807d9504e27ae1b41fba0a2f85c184ef1956f,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1439.9733520268244,1440,exact,canonical-source,profile_8b1d5607f64765be6892d3c5,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,mcp-atlas,MCP-Atlas · Scale,tool-use,afdb5d3cdcc181b462c6c5463cc243043f62f3e34c66a1e08acdd04dcfefee06,d902dceb68d0a6ef178682f48042ea9888c245d9274d6f2a06393bb964cd4f5d,2abb768fc496edc6e0e73ea5e3757b355610979bbf1f39a4c2432ce2a84a1a9d,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,70.3,70.3,exact,canonical-source,profile_671f7a505d0631492846b729,reported,unknown,,unknown,,false,2025-12-18T18:01:43Z,stale,317,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,bfcl-v4,Berkeley Function Calling Leaderboard V4,tool-use,ce2a9ecb4b2ce21ab28a8d84f235cd24a24e419d12e1dc2ffd8eded987ba1652,7f17039421980bb44eedcd74613dbbf5dd2682ba46d9b6e9cb1f9ddc7e9c1083,6a9c99bccbca11748bdf70a2bdd6eb864726c94bf4a029f84ab95898fe96d192,release_7d2560723dba9d549cd10b25,v4-ede5081a24bc,overall,overall_accuracy,percent,true,68.14,68.14,exact,canonical-source,profile_e10ca9a1e04ebb6244b12dd1,reported,unknown,,unknown,,false,2026-08-27T22:30:42Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,bfcl-v4,BFCL V4,https://gorilla.cs.berkeley.edu/leaderboard.html
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,vending-bench-2,Vending-Bench 2,multi-turn-agent,2802a552f3919cf79b50da4d2da202f33225b418ddc3dc670c0fc5983e66d3f2,3fb6d1f6b2a4d183cd039e815557035278badde43938ae02deb41d46329898ba,8c8835ad8aecbf44202e08a057c93f84eca18a816e9e74d6574afe97bfd73116,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,5478.158000000001,5478.16,exact,canonical-source,profile_cc3ed16981114b07863ab4cc,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,gdpval,GDPval,professional-work,b152fe15f701d84650a0298ed88aa3b9fd79af6840de86f372cbf117c723f9ba,f9251d32c4117d4bb6619173b8ef057924f4a90a65af196225fc90c840da2985,2b572160dc5691512f728679f7c793b9cf2eb4c1ed163081ee67c969327580c8,release_730527fa346470703bd19e21,GDPval official leaderboard,expert-comparison-win-rate,win_rate,percent,true,40.3,40.3,exact,canonical-source,profile_ea20236fd014b2198d3a19c8,reported,unknown,,unknown,,false,2026-09-02T09:27:05Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-gdpval,Epoch AI mirror · GDPval,https://epoch.ai/benchmarks/gdpval
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,42cfbe1f2bf5ac74c21877004b45d22fb5e275f9c1238707a72f4f3285615c5b,7c7cb1dd1c896f1b3a91af7417c1e4169bd1828332a39c84c8bb1dc22294401e,6070e5eb3b8465e8c6f5c8390e51640e8e9064c49cf8b5ad9ed333ecc2cba0e6,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,18.04,18.0,exact,canonical-source,profile_adb9d064fb6ec60affbc6752,high,high,true,fixed_effort,,false,2026-03-02,stale,244,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
google/gemini-3-pro,Gemini 3 Pro,Google,closed,48,weirdml,WeirdML,coding,b3771364320f7f4b99a42e35c9cbb91b0d41bb26d077f8871d37b703a04cd5c5,ec1f1b9c1b0b3e6cf1e47c94f1ca6a529150933913854738afaffb1c9b9a05da,823d1839ab557ec28302622ebf9b212fbee470eb44ec550284a1997c2dbb10db,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,69.93,69.9,exact,canonical-source,profile_cd9c62d22639dee9b3c647d7,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,epoch-eci,Epoch Capabilities Index,capability-index,e82c2882f863f6b71ffee6f6c84dc5eb824108dd283bdf8472409f16ddf742ab,3a1417d8aea3972c2b1d549e5b5dd8372295b1fda7cb4ccb4ae990b5c8957e2a,7adf462e707292e2d51ec362ceea354fda4f62e1175fd019adc5fe91a029573d,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,153.22,153.22,exact,canonical-source,profile_4d4e599ec811239ce079646f,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,9d3487cc96aed2d87a07a1c0a63fd84c050fc106b58aa68e0aadd1f0e7776acf,8da61771f4742a7b781a4b75f96976f2a4012ad1348079a29eff731fd226005c,5e55c1514e444ecf4ad66390b0108bf7bc49a54c6267c173b782ff7d6ccf1242,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,27.753889,27.8,exact,canonical-source,profile_626f9b1c4e05a66a41042aa2,reported,unknown,,unknown,,false,2026-10-30T01:23:12Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,arena-text,Arena Text,preference,2abc22484ec60294e9a714258bd96b493a0c977fb35cf4892b19d922d1a4bb8f,397d90a88f911de24333977fc9f63155880c9086fd0f25ac1db23d4431507901,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1476.515607263063,1477,exact,canonical-source,profile_939a9cca21255a0df9b67f4d,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,livebench,LiveBench,general,6fb83191d8d301477d8b03347f28ea129f9148f4f5cf6a874e672c545955a2c1,f385d93b3f327e7168e6d188ecc33f033b1d13d37a76337d7c8d6a75969abf4b,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,71.59078571428572,71.59,exact,canonical-source,profile_ea7550e0cb1aa3e5a0a3c455,reported,unknown,,unknown,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,deb8901d0c21d6a3ce34e41376e012aaf3907e872006bb05ac9ecdf4e94c2bba,22485b6ac0e19bdd1f35ca12afe1fc9b8ea6cf0f8ccf688ea7bd610ffbe9c37c,cdd0be69efb364191bf0190a287662dcbb64c930f0f964558a199ae456ef38fd,release_fa9e908ac04b067eb00182d4,HLE official leaderboard,published-leaderboard,accuracy,percent,true,55.3,55.3,other-supplemental,supplemental-source,profile_87d39e2d4e9d451f3ef67d65,reported,unknown,,unknown,,false,2026-08-26,fresh,67,1,A,1,A,5,C,self-reported,0,self-reported,hle,Humanity's Last Exam,https://huggingface.co/api/datasets/cais/hle/leaderboard?limit=200
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,7ceb053021ca94c49f4ab34986b453979e748d288afca21d65ff91fd8ab2fc5f,11e28dc791bed57fa4c2c2b246b231403ede2517252d4a6d27e2ee41b65b118f,13ee860e88f983adc2c47da369567d12a1d1f8395723e394cb1382efe3f68f5f,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,39.851715,39.9,exact,canonical-source,profile_626f9b1c4e05a66a41042aa2,reported,unknown,,unknown,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,gpqa-diamond,GPQA Diamond,science,69dd7fd206dd82e2099d3a20a7ddebabed028de8e50f6d6206ac5ccb303391c2,2ef1dda6ada50dc3a861a329521fdf7171326e9a4b4813aa37176f0fc1805c85,ba0acd2b6380f6d31223b2c703e121207197f3c3baa061cd4b93ea020e111f23,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,86.364,86.36,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-08-31T18:49:02Z,fresh,61,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,mmlu-pro,MMLU-Pro,knowledge,74bd3b9deaa56bc2f5d898fd273861f391a26c9ef97314f67c247937e6f51645,79d24c81778865c06f24e0d3975da20fd8f554d4f2b6466d267d3d761742d626,6fbfa8b1534aab5b541593f750003c3cc5fc52b021490b23a4a333d020c7debf,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.059,86.1,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,0618903d95a8d584d9390b3cf584dda512c42555bba835215d194156eb16270d,7cb500e95308f79a05b356e35f8df30297ac5359065c486558dc820e8a591309,0906296fc2e811218f86a4ba9b8dbcdfbc466389c3fa10d3c7bfda2fafb8e2f4,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,93.88888888888889,93.9,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-08-26T22:26:55.000Z,fresh,66,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,frontiermath,FrontierMath T1–3,mathematics,d2640f63083ed7df8aa9f62943ae4e9091a3083e33913ac2e34c1efb14fe9541,a2be6f3f648ded04ba25e512794af6029790cc28cbd7dda8cf1a2b86643189e8,38abbc3adc3f7b833dbc0883b188a1ffc38ddc878e9e36dff4d67f43e808a1af,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,55.78947368421052,55.8,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-08-27T13:29:52.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,arc-agi-2,ARC-AGI-2,abstraction,f095cb78e5aa194d2430fc8c557b2da9d42cd7a2b81adf0a227b3554e4c30e2d,5a313256a133813e76fb78d71ce82ab0083b16d847b9908d19a0a498427991ba,43a790acf37c517ffe418e9a2ce33e021e3dd50d977f332795071689b4f024d7,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,65.83333333333333,66,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-09-30T18:53:20Z,fresh,31,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,frontiermath-tier-4,FrontierMath T4,mathematics,53557551c330f69fbf2cc13d901dcee3989858354395db04cda31c05e3e30ad8,d535456016ca2fd782e6e164b195227bf2b7de6b8ce8e2a8bd103190428d7abb,f8aa604bd1c0f91a30ac21123b2b364ca02105a735bd72c50c5f559b6a2d2ed3,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,17.073170731707318,17.1,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-08-27T13:29:52.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,mmmu-pro,MMMU-Pro,multimodal,b30238cb6fa78f5d4a1b4848b13d09134845ecf35051232681ba360c30b900e0,1cec2b4b5e1533803cd47356c620d87460350413fbe64ea2c46d168e14f3bb2f,10722aad1392375e0e127a11421484e6486649b1bc97eb8f4d94df6e1c672468,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,86.012,86.0,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,critpt,CritPt,science,70f68ab91f20c4c66062c49d4a4d3a774526c78eca795ef85947f5c11fd07e78,ecac7e2fc0297528f621077244468d4a25e5fcde5dfbad33dfd9b462d117d2ab,cda300ed2945c80bbe368789b74d06a20a060a0ecff520d376230e8f02e0de36,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,15.4285714285714,15.4,exact,canonical-source,profile_d32f80db8f8edf3a161d1982,reported,unknown,,unknown,,false,2026-09-02T21:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,chess-puzzles,Chess Puzzles,games,2d55200bfebb70a9db6b40a92435a3f895b394fbcb6bf080a8e379622478d934,3c0cf0a317fa4d773377bb16fa79663c392de3122edb50f1e35ede62c45c5bad,b4e5d8f5043c440de4c94daf32e4fc02f32d603d698499cbebc8bf50f9659a0c,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,21.0,21.0,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-10-13T15:54:34.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,arc-agi-1,ARC-AGI-1,abstraction,7900244bedddc2c363145ec01711d601a6ecc0dceae929895b28ae599869be4a,d73e239b65ff09cf47179f4e2fcf9d6717fd010b4bdbf64081a8414ff909d5e4,983238485ae334c26bf791458ea3f413b5ef0c8c1f03258aebbba99789675206,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,91.0,91.0,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-10-08T22:23:14Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,swe-bench-verified,SWE-bench Verified,coding-agent,c55968d4289ba43277c4605c6bdeedf1079cf88c4626798b29d8c84ef2e47431,fadd0d0cb7b3c181614e1ea8769502249ebb299ae1d3b934f72db024b14d66c6,eaaef31849ec7cec8b278a82aa41a3cf59c6c7b2ebfe351ba73dc096c1b2f259,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,92.0,92.0,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-08-28T00:52:15Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,deepswe,DeepSWE,coding-agent,cb923b3372dc619c5d97dd0cace2f1ab6e2ed79fc013843aef22c7e82e7a0568,23a44ecb25a1383a9d24b3ad23ce046ef78317b1f2e423d712186d7a1f15d39f,600d6a11dd0c7650a51ed4c33e6676ad6579879e8b378aea81b58fa00e0ae39c,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,63.39285714285714,63.4,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,cursorbench,CursorBench,coding-agent,5de983a2213ae8da2024a1892d08a58888d5a7dc26188bfce6cde8f5360e0769,c6bbfb8a96e741a34486573447112c5c0aa41af123810d1d0c51193f40247123,67f9f9b40e690cdc0b038d5009351228093a756098f53fcfee78843cf140a6a0,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,36.8,36.8,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-10-08T22:23:11Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,frontiercode,FrontierCode 1.1 · Main,coding-agent,bc1e42a6bb609eb1578b20b9ae3825eac1e23d69a6dfe63f68d05d02fafa2a19,4cf9e64c8296f58a06d6265a5135e7a572488a2ac73e423a73b4d190647ec9b7,8cb81a03ff70259cef6db7966956279ca9dc3fd8f754cc3aa01719a706d65d7b,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,31.830000000000002,31.8,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-10-28T21:25:01Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,ioi,IOI,coding,4fecb7b5fe96d28f44477d00eea3ee39a56abb09efc66d3107f2736b09082100,b29830093ba1074c77a4e217bb74553fa83789de1da478a4ebc1ffdaf7676451,89c8f336bcbc552acb7fb1ab594d4c5eae7ba289d39321f7da95665d4151de35,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,52.5,52.5,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,livecodebench,LiveCodeBench,coding,97cd5d613dedc19d2984979a77cabdd521a9509b101b1c146a9d322cf9a5d920,54cea8f77b707fecfcd51d5407c42d021fba10ef0903d6aa3f29ed82fcae9339,7e2098411bf4d25a7549ce193a1a70c35cbeaf4d6229edbc8c3ce65cf4acb865,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,80.51,80.5,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,scicode,SciCode,coding,89e477a09683725297bcbacd7366b588acb6781ed80a72c1027a0b4b5cd694c2,310321cc916493bfdc993ed57a7814593ce8cee66d28edce66044d3c8546d829,f18af38465f37df2ccbedbe4e44a9262b759b5e6705605a37f58b3188cb7020c,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,51.6203703703704,51.6,exact,canonical-source,profile_7f5a892f8608b1bcfaf9c196,reported,unknown,,unknown,,false,2026-10-28T21:25:02Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,ale-bench,ALE-Bench,coding,bb6e82acc810b2e1eefe95e267bab2be09116d8e581170c89ac1b46f0ba61695,06f4b3f0e704c566e5cb1bb38ff933962905c5e6ee3c62b7413a1db61f29f492,871e7a4ce82d510c855fb65d3bb95b88e7fc08a92581aab4e2f537171b51df37,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,303.55,303.55,exact,canonical-source,profile_549ab507d1f6fe5b6e3cb294,high,high,true,fixed_effort,,false,2026-09-02T21:27:35Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,vibe-code-bench,Vibe Code Bench,coding-agent,d3c922ed9af924baf253054d61ee838d5e97d57c6eac64ce7d2d065e7a9bd98c,f3ae63fa468bd32f5890a6a2e51efb2db2ff611cce74a6d0286ba8aa667066bc,fc496087d9740c9dd19f058a520713a048f21283f2524025ebeb98759e7438d6,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,30.759,30.8,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,webdev-arena,Text Arena · Coding,coding,fbcc424cbf0b031c3c61be61d8f2ad505012137bfa678e627718f148c498f4a7,9cffc18df28a81eef6a3feaa20b1c8ce834ef0a5bd17a6da259333f4dfe19f8b,0f8e950b9cf98b2ce00266645ac047eefa3fea4823a3a40b09d6e44c4b0668b3,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1606.6486731495352,1607,exact,canonical-source,profile_d35421dc8bde7a8d07e82ffb,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,programbench,ProgramBench,coding-agent,2c955b72c37574c334f9f87fecceca50b20737887bdbd97290be0c9dc5532819,5f8ab151b62c4767ad29454fc0c0e725ede96bd618f9cf87431b463dfd79b030,c3ff61eb4b72330d8e0f03a14ae1d0d4bb285fe548c6627e75b1e25d141a02a2,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,08682204c130ec84e22c2fab61698b619eec52b38961d52b19cd0ff109ff1ede,351c25277cedcc5f4d55bffafc4ba87519c3756908993afe5aea66149f69eec1,257b32c7a4567a3f859c0e3d0dd78686e4e45b052e7fac2e6e540f245c84eeb0,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,62.921,62.9,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,6da1796139959b91feb92ebca7ef2cbafd80dc6dd5fcff9f32baf78d76375e2c,cd4388d9f0c377a3d58ddf62d13164a623c5fb4494acbb0fa1c4eebeea121760,c8f224bf5c0a14a85fa72a49eaa756c39cd9366b8ae75222099568309784684a,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,37.374,37.4,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,sre-bench,SRE Bench,terminal-agent,e0ef327efedd37a8368dc09eef029bc438f166c2ec453e3d0a4262b025c36930,e480326db71798dc1adc5e7f62a26df416be5d97805999717fa71542fdd8d2c6,b1b49e04ca5b9f0a65e383c32a5cb751f9b877d63d20ae00ab459c2d1789e6e8,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,0.763,0.8,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,7cf6e9201f688ec9ed4df9e0bdc9882c242f49237d015308b6ef4aab95cbdd90,7f3369652346fad3116e5d1fc5baba8183dc234515b82af66f35f4171c54d756,f394d968a83273ebd001decf638fa6dbc38319aec1849a6a23d3a8449fbd6409,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,47.216495,47.2,exact,canonical-source,profile_626f9b1c4e05a66a41042aa2,reported,unknown,,unknown,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,apex-agents,APEX-Agents,tool-use,8e24f2964f876e540c0747a2587ca6735a6fe4976615ba5e4ce2027a4f6d5f12,16068ed862433ee6b3f760755d376e1e4fae65af9bab18a6e1539548de04a0fe,748f9e1e1c80f60ea6a805a41ca001a50b989b3ac76a07b81ff5ae423b61bc34,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,52.800000000000004,52.8,exact,canonical-source,profile_605f0eded166b30ff74e07bd,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,arena-agent,LMArena Agent Arena,multi-turn-agent,4f6beb641cad092d9284a690cd28bc2cd043968ca73a5a36e45622505df8806d,79a68fcba02ead526213d6a5db73c90208055d895d683f34236bf42d5185f8e1,f32ae22a180251071c523302d889eb0a7b9cedf0fc80ce48c2158449d34bda72,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.023262807029308576,'-0.0233,exact,source-specific,profile_9b74d27daa9d9ff35df369b7,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,a80a3c34b2cb6326e9f439d4cbaa7f1ad50953e3b189dda8dc2738e4966686f0,3e2c5fb74da9fa3d9b57097661eceaa49428714ea955869c0be90d4fa09e8657,d2d1902c6ed889929fb7ead8645016397d6e778d104975fc8296ced9a9393a71,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1649.37,1649,exact,canonical-source,profile_626f9b1c4e05a66a41042aa2,reported,unknown,,unknown,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
zai/glm-5-3-flash,GLM 5.3 Flash,Z.ai,open,49,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,1febbe40b94b58e448bf6d3146d73e71b332a18e52470d143983f88acdf49300,cee37d1440e5e3d3a09140c1f2f4e544ba562a2e1ee18c6d7414fa3759d1d2ae,1726706cba26acaecc716e2891cb6e6729cd38f8e0f730d847be899d1d50f1b0,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,2.857,2.9,exact,canonical-source,profile_d3256297c99cce45effb308c,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,epoch-eci,Epoch Capabilities Index,capability-index,3da7b3b6c5a3a5b90b5db610d1881e44b52ea9ce80d4799b303aed31214bdba2,968f700d8e840e207e299ee534b61f86a1955ee4252fa61ec1c1ef51a89b51c2,980f8f93595b05b888c932ddc39ee9695a3bcbe72f2c039fb66705210ce2b2f7,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,152.9,152.90,exact,canonical-source,profile_da95b541a68b4f050b9475da,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,arena-text,Arena Text,preference,eba54b3399f0a45b5ffe6a58662f9efe9ad83489ae5e16a26e9358a1946b9ee7,48f7035713a59a0af997563bed8c8bef8ff9aea04b06cd9676ac973ffd3965be,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1471.9506765539006,1472,exact,canonical-source,profile_5f4b5142e38dcec1edde2831,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,livebench,LiveBench,general,a4d9a46c15df94831042edeee6f8b891198849d05a0b713aa3c656e73df8ca24,ec4361f63ef16f39914180cc5b0e517f7cd8133e6d2c84f7e8ce48879ca47cb6,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,72.99002380952382,72.99,exact,canonical-source,profile_02f18483087dad439956139b,medium,medium,true,fixed_effort,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,ff5469580f4c8a133ad1cde008b35a24d9aa88e1ea5e12550f416c0053755308,45041b8488488b3f51a0825fb39be6084ad10a2ce27b301d8c35e49eb1377cfa,f3369573ec5fa708c2fb1887738ca58a6da3b6f6fdd4087151c44904e7e0c723,release_c1afb6c54b8ceec8bc1f121c,HLE original full multimodal,full-multimodal-2500-no-tools,accuracy,percent,true,34.6,34.6,other-supplemental,supplemental-source,profile_1a899ff2f2c99f660795bf87,reported,unknown,true,adaptive,,false,2026-06-30,fresh,124,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,anthropic-sonnet-5-hle,Anthropic · Claude Sonnet 5 System Card,https://www-cdn.anthropic.com/283ef97c476cf442c91d9a37d5b214242a55bb92/Claude%20Sonnet%205%20System%20Card.pdf
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,7038dde0eb123fa9bcae759f7cce57b8cbc48d2dd838b871b2fd17954b6196ed,d8098d508d4db58dc96463124829d48b7c35450809c3bac6d8031e0ab3c50235,b982a6b744e7fd55a5a61ca3b7e90f01a548066ccee70340d09bdca8af5787fa,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,11.214087,11.2,exact,canonical-source,profile_e0bdc98a9afa7f9145f07deb,off,off,false,none,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,simpleqa-verified,SimpleQA Verified,factuality,5be1d687e66076e67e980a89293ef262a5f049b95a195480d33f1b20ea81c63c,08be4d2e546c4654e4297ad61f34549ba722e68740c4fb04c299ec4c175a70f0,b0e6d3b8121dc6f9eb2cae0167640e4069ac1148c0d8cacb2a79d747591e72ee,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,32.800000000000004,32.8,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-08-27T19:27:37.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,gpqa-diamond,GPQA Diamond,science,3387971a9211e952e23148f51ae38d8e6115f3fe1c1bf3399647f9cfc084acb0,1b1fec15bae678cdd6ec970b54d94f111d30955fedef1e428bbb05a5836fee72,9823003a075d79bbbc67d654070bf256fcbd99aa52e3e41701ada9dff0700b04,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,85.606,85.61,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,mmlu-pro,MMLU-Pro,knowledge,6f655341c19f9e7419b7a3043c2bb2353dc8e88b58f5dea29ca3625b117fbf92,d7f91be7380a159287722ad3b73e57d0b969cf12796aa44c884c3830a3f2bf96,3cb39cd3839198940da29f05c4c5f7967c69f284104e7cfb8033c3968b9c1d5f,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,87.341,87.3,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,3811652be3b0fe80aeeaba391b48220f97fb8af4478be65b69c7825a8629e3a0,a4f378820f42bcf8b92239a8aea8e141cad014b52dced10da3690d51b746a189,0f46c770def16a961a1818eaf4b929b42f199ee5e3b9864469ec01a2413995cd,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,71.11111111111111,71.1,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-08-06T22:21:16.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,arc-agi-2,ARC-AGI-2,abstraction,4231a538322dfccf09c8ae1601c6a8035d5239cd383ce76bfb9153d9207652b4,a67d92ae67f71abb062495d8ff1c179bd7266130c6d5d8f75b37fab002a994be,11a379dad5a9bd83ac82b0619340f7e3af4d02aada4a2643ff8f4053b46d0328,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,58.330000000000005,58,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,mmmu-pro,MMMU-Pro,multimodal,ed8343c16b1addc92d65ed9da2e909575f00445b3f163432ad58b88d57bbf31e,913c8b928f388de356b7fcd1f1ecf22b8431e81cb43c39ba0751a19c4e768d5b,aa27cc7123f0eb619196124a5c8b8020d1af2c5faaf52acfc63a1cdb5bb3c67a,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,83.584,83.6,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,critpt,CritPt,science,b56c919d31a2aa225395acdce9f382649c772e07598c6e47d0a925be4b5e3c55,9bc907a4d2f34dc58209f40923b9c18b12a8953c9ae5b76c341c6fa690fb7bf8,a5b147e0e3ace6b42d6a12d624727187fa8b751251905e2567ddc62efd7d79c0,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,0.857143,0.9,exact,canonical-source,profile_e0bdc98a9afa7f9145f07deb,off,off,false,none,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,chess-puzzles,Chess Puzzles,games,a04336f1d7cf355a900c576ac9cc982f60458f8685ecaa465bba8ddbebf7f033,822548018c46cc69aa86623673db241dacabee2e0d754a28dbcacfd7d8da4379,c66da0092512f74de68fd680771d6018e5452bf64c86abc4105294326535b87b,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,23.0,23.0,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-10-13T13:44:15.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,arc-agi-1,ARC-AGI-1,abstraction,b2cdb1d5bac86a9ff0b2bf4000dd98cab2a0b5519dd508ebaab4253dfe6a48ee,4b99b944cf57eb25a115c48d921017a24927c2daa7bc87ea6609816ca1f37255,1524582ed693ccee7f08afa2bb1f91e502bc548f5f232eddfb3f196bda91939c,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,86.0,86.0,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,swe-bench-verified,SWE-bench Verified,coding-agent,ac475b4d74b4b8dc67abbaca4bb08ddef41781ca3096731376a6a9087b441dc4,c2d46aed4af0a99d2ed5858e99a4a1feec24bf56ba04f76bc046ecc9fecdc9cc,8e5f67404b253591e5c15671113eca6c244d022802e49f968e736452412c83f9,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,77.4,77.4,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,swe-rebench-v2,SWE-rebench v2,coding-agent,8d5c6403f38b12208106d7a4460c0b870b1caede8d7fcd6ecc9539cc5229104b,d5bd5ba287f21671b93a9c7a9d584cdccc472474131a2a2b0f115faf67c5facc,690f71bd7ea1ef47ffbe550903858995f642b0efe28b1a7c4ea90de3356334b9,release_72d764965dd7baf34f983249,rolling-v2,2026-03-01/2026-05-15,resolved_rate,percent,true,51.272727272727266,51.27,other-supplemental,supplemental-prior-window,profile_e6ae4e8ef0b5692fde876ef0,reported,unknown,,unknown,,false,2026-05-15T00:00:00Z,fresh,170,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,deepswe,DeepSWE,coding-agent,165d363bb652874fea00c9e143c951ca2e1a6450a3fa20407b50794f7b4fd176,da795be427acf83b346f70b1f2dbb787623e71a42c3e8aeec4627c3bdaa176e9,ada28153a47df2486f049859a8052f5c5e8be46e72fd73892686844c49cb15d6,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,29.933481152993345,29.9,exact,canonical-source,profile_d4a7e98301a45fe263315cae,high,high,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,frontiercode,FrontierCode 1.1 · Main,coding-agent,f7de25978d229d50f1f62781804744fc458fcf7b85e51a593b9a5e084d708cce,46690b72cc6c4769a826e86b5058060c40886b3e2f5f7c6d28a8bf153b490e59,363f5ff210e51b2137fdcc9eb345984fc28f27ee9565cbde91a4859553167d2f,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,24.310000000000002,24.3,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,livecodebench,LiveCodeBench,coding,61c6ded338265d644829ea59503d8772387ccc977dc031e72903e9cf5d89331e,871aed967fa02ba434b002838fb7c593796b2d57278848e5569743d6aa67c61b,aad869c531a6e7a41ee6b771fa40c013448dfd0b1a25c1a825727394cdd11df4,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,82.091,82.1,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,scicode,SciCode,coding,b651a3be322381977b8620ed2273f737abde6ff69a16a96964acea8364e85daf,472dd3d75969c41764766376ae2bd714c60b35227632f4dc43e58eb5d13a7ec8,efb8fe7fec2a2450eac8fa367c73e99ae36b1f2644c728fa4cad58d2a1ab1c4f,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,46.7592592592593,46.8,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,ale-bench,ALE-Bench,coding,b32bf193a1f48e0fe7fa63322b8796d84f36f6851d14530baaa52fb1abae3e05,5711530b200dc72bfbe0d71e9c8e8ed3ab023cfd8c97bba5958d6c73cfcceaff,ca66496b8f55b81a101ae4184d534d5bfc193662ff3fab9d6667fef5f94674e5,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1327.3,1327.30,exact,canonical-source,profile_02f18483087dad439956139b,medium,medium,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,vibe-code-bench,Vibe Code Bench,coding-agent,3ba00c5a446aa05b3807c5aa6d8efc2d2025854a8fae8bff21e2d70974fa221e,f1572b7f5df743ce5ea86c78572363f7f1f67f4f30751244b2ee65dc38cec2e6,dbee190ae846a89beae131211a82cdf080cad285932d0350b9fde2c272a22d1a,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,51.476,51.5,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,webdev-arena,Text Arena · Coding,coding,7c4f0ce11daaf748bc508f92e1816202aa1b910443a128e28e187e33b282288f,1a458ac971e48b351480174ab0279c01ca592185e88bcf615126059877fb4c99,802905efecee23aaa096cbfca9dfed1ec427d2d6e0bc76b1017d71ed61c7d305,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1522.6524947374642,1523,exact,canonical-source,profile_fdb9efd11c7d6668480eb672,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,programbench,ProgramBench,coding-agent,c26c9b6c5a94ee20074df3ddaf898111e17b0ba6f661020468211d047b93d58c,b814b9b1ef8f151a38cac64809ad1cf105f56e92535bca40d47d367bdfd647a2,52f7c8e9d70bb8553974275905dd71dbe6ce063254ab9a26da54298ca112d509,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.5,0.5,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,7b3f63fb2ac19f8d49fdb24412102cf6f079e887498d5364285522eb60a3480a,228f77432156606581145f4b4723a4cb32e8b16dbf9addcce9cfbabfb2c95fd7,8ed890fcf8d1451ae67246ce2f9adab263633f997f49cb98fc511069ce87215a,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,57.303,57.3,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,mcp-atlas,MCP-Atlas · Scale,tool-use,73df22b3450c48d4f2b5036e90e24cc71deedac3173dd36d723cedfcbf246dfb,b1d78a5104618d2fd4e4338eda46f80f5175ec8caa07fcfce7be5bf90f951262,0d2bd69aa96fd9b8723c539be0a71ba1174c8293dfcf4dd5f66b18d300fea3b1,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,69.5,69.5,exact,canonical-source,profile_ea6c6d9a53bd19b358288011,reported,unknown,,unknown,,false,2025-11-06T22:29:16Z,stale,359,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,skillsbench,SkillsBench v1.1,tool-use,31bcbd09a0b42b64548d2815acf867e65b21cd4ecc5e9efd0b00e05e6069f9bf,209061e382ac90e2297519b0e9919b8330faf6c64a91d83a31713a0db3e20f4b,c2ed2025e534826e5e627446f4493a128269710491b35eb2122abbaf19d56304,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,47.2,47.2,exact,canonical-source,profile_f443077f18b018dfdf0194f2,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,apex-agents,APEX-Agents,tool-use,f74ef540e5c350970961b4f935fd3da46af26eaa0b0a44f2226e1ee1220491ed,453274bca5a7a2d5589423d36f84c333a5161e0e59a06b6dccca580924cfc6b3,f55b74255d42e0de1f4b986a9e4804117c4d727b32e7720efb953f25c62a8c39,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,43.0,43.0,exact,canonical-source,profile_d4a7e98301a45fe263315cae,high,high,true,fixed_effort,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,appworld,AppWorld,tool-use,97c558a666620ee86562bba0f0c3f218f73282a75af6151e219cb8ba46bf6cfb,6336f94d4b9c028c2bcb250b8e0efa3899b18f7da3d69c822fc82f50f5e76276,fffd4ca8543c5af8b9de05f6bc02595ac6a01c285f093c90520fc13270ab41c6,release_3fa145932216c0afcdd9dae6,official-c1f56015cf7c,test-challenge-all,task_goal_completion,percent,true,98.3,98.3,exact,source-specific,profile_c0a0659ba50bddf183888efa,reported,unknown,,unknown,,false,2026-05-21,fresh,164,0,A+,0,A+,0,A+,official-leaderboard,0,official-leaderboard,appworld,AppWorld official leaderboard,https://raw.githubusercontent.com/StonyBrookNLP/appworld-leaderboard/c1f56015cf7c3441ff1933f5bfbec879798d7bbe/experiments/outputs/_leaderboard.json
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,vending-bench-2,Vending-Bench 2,multi-turn-agent,ef1040d9f5d5c4da62a23a33e532859bdb4096adeef9e21721ed8383a55942e2,e55c98a6615e13f3183066b69bc5d1b10c15378771758a182325453d145f993a,592565df30e1bc460e6d17b003278a36e3f11fe81435287d81c0fdf1680364be,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,7204.139999999999,7204.14,exact,canonical-source,profile_2d36a21873281878c5a3aed5,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,b00ba0b8f87159c55d7bb32e98c9f23dd8afb4bc817307e304b5f38085b82733,6a9189e1f9b59f1c2329b6a7d58410430c63dee837d08e7b852a2c3378a545d2,b597932890c8fc89967f508f21a13feb75690351602b3d29d5070926c8463835,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1240.5,1240,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,osworld-2,OSWorld 2.0,computer-use-agent,c297b4714f322c5e78a6770bd01b28c5becd745efe97eefc230bd777881a0051,eee6d26c91842eebe82db7a4092b9e10eb9124ce3597e7e411589530db7c2b13,420f7d7c5f6431f39a1c12e26e98cb549477c9ac8537259a07c290e052f5f225,release_15c5525fe39f33bc181dfb69,OSWorld 2.0 current leaderboard,binary-accuracy-500-step-budget,binary_accuracy,percent,true,8.3,8.3,exact,canonical-source,profile_258fcb8ee8e5667bc1ac3d85,max,max,true,fixed_effort,,false,2026-09-02T09:27:06Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-osworld-2,Epoch AI mirror · OSWorld 2.0,https://epoch.ai/benchmarks/osworld-2
anthropic/claude-sonnet-4-6,Claude Sonnet 4.6,Anthropic,closed,50,weirdml,WeirdML,coding,b7d5a83a4b3d3087051b276d3d55b5bcdb65c2bf9ffe50c40d12364c3e0593ca,4c30244f6e2c4dabd9cefa278cb5212af4f4a531821adb61deb172ad8a933434,88463472ba6da2c04df6df5156dfed5441a09f74efd6bddf4e43c096e3295fc0,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,66.07,66.1,exact,canonical-source,profile_02f18483087dad439956139b,medium,medium,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
xai/grok-4-20,Grok 4.20,xAI,closed,51,epoch-eci,Epoch Capabilities Index,capability-index,2b82c6bb601c2f5ff408933a73794372abe7224088fa2cfc761c32c0c889a768,e6c693a7330866c32d5ac7d9aa6865baa16da99629c3f3f074f0ea8430567c92,a1ccfca62b1f49597d8087fec9f2cef23768e4b846b52e21eb4a41ce8b793c17,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,152.26,152.26,exact,canonical-source,profile_907345bd817b4bd0177c506c,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
xai/grok-4-20,Grok 4.20,xAI,closed,51,arena-text,Arena Text,preference,6c94d0de4ee6551417683ecc0f76640e39c9c4db5bd4c9a8ac14aed32bb7958f,351135dae2f83c5ad92d6c9e4c14aa68f02ee04d29ca7cebbaf7c901d03393a4,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1471.7400417518265,1472,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
xai/grok-4-20,Grok 4.20,xAI,closed,51,simpleqa-verified,SimpleQA Verified,factuality,6e913b1f44638151bc6336a71298e51db43904fc27995b1eeb31e637c3fd6919,aab135ec8b860538d1e149125c3ac67615b2dfe03f1f6788eda573b392a3c222,4605760acc55f59e9cd7de869182ff241c553644767751f7e85e6c82330175d4,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,30.2,30.2,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-08-27T19:29:44.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
xai/grok-4-20,Grok 4.20,xAI,closed,51,gpqa-diamond,GPQA Diamond,science,37b0c4e054dd8d9b781bc891a9e7021248eaad094a3c77c9bd56e403f8afb39d,525b2e379e2a2c881901fdee3bb659b21a2911e49352b7ab7b4f16e246845ab2,d139806585c7948dae97cf14f6c8e3161b21a91323f1bdba4dda4c953cef256a,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,88.636,88.64,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-03-13,stale,233,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
xai/grok-4-20,Grok 4.20,xAI,closed,51,mmlu-pro,MMLU-Pro,knowledge,37e42b6fccaebf48ca4662816ed6942ae4f72e47bac342188b0e4ad8a9e4f6e2,fc6c6e7514edbbd32af7ecfd9d31c60b8d89d10f0d3efdbc855ec7cc930769e2,6670a4d34719b13c91512ccd0718df885c2b66d4e4819df6701967bf1fbf6dfb,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.254,86.3,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
xai/grok-4-20,Grok 4.20,xAI,closed,51,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,dbc47bcc2ff7de4725ca2c16341a9ece1efe1fdf708281b53d7fe43cfb6cb3f8,bb48b7e5cec6f16ac88e738f880146017f03f25224a07feb31e2f051d2507e73,82e13eaddc2ac54e19768a7bd6791e3aa6e226b89a4a8d14629617a9d47cd1a4,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,92.22222222222223,92.2,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-07-13T16:45:48.000Z,fresh,110,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
xai/grok-4-20,Grok 4.20,xAI,closed,51,frontiermath,FrontierMath T1–3,mathematics,32ba05a67afa0bc857f3ed8724713c0ea448f6b9250c077586fe136f105d909b,60918ddf07707c67d5cebea0f0997eee0657f32676d342f1ab03929612d19aeb,107b35476888ddc67d09c05c0c48dd53aee5142afadbc44eb185acc3f8cec81d,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,44.91228070175438,44.9,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-07-13T15:51:35.000Z,fresh,110,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
xai/grok-4-20,Grok 4.20,xAI,closed,51,arc-agi-2,ARC-AGI-2,abstraction,49073a92e6a13982bad282b7b6efdb6971050a620f80933f1879318bb50b0ed4,33d6687ead19ba47d8616ec109f4543df42fe7fb2684ba4c160d01c92a6144a2,c3fa0592b6441af3735d852e2d062d733783e4f6e00f6d83561e492ebf372c72,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,65.14,65,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
xai/grok-4-20,Grok 4.20,xAI,closed,51,frontiermath-tier-4,FrontierMath T4,mathematics,acf13b329cc9fee84f57f1616208fd5731eb697b4a3918f74783db6d7b8ffc81,ccea13b7b9f513315847d29ab4a00e81d36fc121c3cc4db162a6ff33c55c0797,55c3e2eabbee1ba0c01a0bdac6bfb587021b6787918995cf0ae59abb2103760a,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,17.073170731707318,17.1,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-07-13T15:51:35.000Z,fresh,110,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
xai/grok-4-20,Grok 4.20,xAI,closed,51,mmmu-pro,MMMU-Pro,multimodal,fe7318023d1b681255d990634c27d4fe8302b16cc742d72a916a8dac13eaaaf0,f2cae986e2b6c4d182ef30afa93926a5d872c4346bac471125999dea792f70c5,baf76148bf8f3e45d14335dbad417dfdcca6a6314a043d95db1f64e234bbe891,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,83.468,83.5,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-03-13,stale,233,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
xai/grok-4-20,Grok 4.20,xAI,closed,51,chess-puzzles,Chess Puzzles,games,6edf63d19ca91f2d5bb5d83773cb4d232ba03eb5e8b25427e99359da642ddf0c,76a2b11fd202f219d49061f4eeede64917166b27d7ccb77a8fdc736d3f221114,4f5dace898c3c357370c69540392f8287ef657acaa458f53cacfeafb48ae6024,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,28.999999999999996,29.0,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-10-13T13:34:38.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
xai/grok-4-20,Grok 4.20,xAI,closed,51,arc-agi-1,ARC-AGI-1,abstraction,13df1fd4e0d4f61d2b5ba15e71860dd34c8ec131e14cc685cae19a66e83878b6,77d2d19ab79c543cfe1e321998afe6ebc2f98bb4437ff13b03bb3285edc150b3,c6afb855bbe79a876e8648668d61283fd4fa5aec6eb157ac1907db4cb7ba6129,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,89.5,89.5,exact,canonical-source,profile_899434b1ccae0f0aa50272f9,reported,unknown,,unknown,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
xai/grok-4-20,Grok 4.20,xAI,closed,51,swe-bench-verified,SWE-bench Verified,coding-agent,59a6877c6755585e1cb47a40a6de6a87400dbd530797baf45289ce9315b36807,3bbf8edf6eeac9833a624a1721cb723873b01aa6ec69c7d836a9de21a21f61c3,cf0baf539458fa193ad58ce64c575c7887469b4195c3ed1940d846835a7d72bc,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,72.2,72.2,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
xai/grok-4-20,Grok 4.20,xAI,closed,51,livecodebench,LiveCodeBench,coding,40c0ce67c0b095d43170b0bf1931bfb0a7a7aa6e7c199af25ce856feacc34d24,c4cd50e107e56f2ec7ad6ded3b2fa557f78ce827f735b84dc04c0d78394d5013,c9ce2cd99fabe1fd928be2d8007501cf70f10e002dd25eaab7a4b664aa20afcb,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,84.265,84.3,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
xai/grok-4-20,Grok 4.20,xAI,closed,51,ale-bench,ALE-Bench,coding,4312ece7fc314c2863a7d980c30e4d7fbdb7b55c30c38c68f94e110b2e5eb320,e11a42d03c713212717b1a4894aba20cb28ab220a3ccf013088bd5ce09a1305c,8b0eb079783f5c9bb32287042ce567a1452c3399a3ddbc9506476318f495e6ac,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1150.28,1150.28,exact,canonical-source,profile_c7ba76e47c689d327667a5a8,reported,unknown,,unknown,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
xai/grok-4-20,Grok 4.20,xAI,closed,51,vibe-code-bench,Vibe Code Bench,coding-agent,375e89879f26626043f2e5123ff159f04b63159565f6d1a75c729cb981d9004d,d1ca00307c37674d1dda7ac5a80a4a65fcffb18d4fd5db6ff3002e7c2e3f1fc4,4b84df3c29186774ea86cf3521d8d64f489b089eefabe0935bd4420b60916c85,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,4.063,4.1,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
xai/grok-4-20,Grok 4.20,xAI,closed,51,webdev-arena,Text Arena · Coding,coding,6854a66ecb1bc9a23a345d8f48cf0840928239511d6f72df632ead68cef3c501,1cd8ff0a8f44ab75d69660c552013be27a09084dca7a7fd307e2dabe324dc2f4,e640a3952747a7ed76657d5adbeea1fc2a75949a0fc4a4404f765f97ef071e2a,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1375.4624441703345,1375,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
xai/grok-4-20,Grok 4.20,xAI,closed,51,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,70454c1f1d47da3efc2702ad47f0dd903078b56c5d5123e85704cdce017a4e29,0a99b96fa3771a4202757a3d5ab6de653e5e2accddddc5bbf80d3cb129e6c7cd,16472b06cbcbcd02677814e45388df9d6d7b2063dc0912d53d664b202a63ae35,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,44.195,44.2,exact,canonical-source,profile_586ef081fe8e74ae64d06ed8,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
xai/grok-4-20,Grok 4.20,xAI,closed,51,vending-bench-2,Vending-Bench 2,multi-turn-agent,af9ed87e23073ed45b82789027b9b65a14da4390bebdbd0472981a1c698a56a9,9f95c10ee1e0e88c20da99fe61dbe4a5ae6567761994ad450bf4f2cbf34899aa,9c035e4c8ddd029fc4d331442cb9952ff1acef4af6cb436912b4d82ed4942647,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,4662.846666666667,4662.85,exact,canonical-source,profile_91997adfb8f527350e6f7228,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
xai/grok-4-20,Grok 4.20,xAI,closed,51,weirdml,WeirdML,coding,7d59dcc39743128197e63a95fbcbf6bb3edaecc73c1dadb7c74608ac205b7a11,33b2adbac8ea65752be7acf4777db39d6d3361db8dcf538dd3567ab83abbc7b0,fba795302480298a60aef10f1c2059807bfae19ad21f591caad5eeb0d0cf6c22,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,52.26,52.3,exact,canonical-source,profile_97752946e06fcd42f05804fd,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
meta/muse-spark,Muse Spark,Meta,closed,52,epoch-eci,Epoch Capabilities Index,capability-index,1cf25c20e0d8fd4751b909db434615cda514a2c39159f0adb97f2b9fe4cd824f,acef1190bbb1a27372516d5e874b38e5636675f4528ef7c7b786ed0a843799db,539c531a6d50f29417df7c7cf89ce26576f029275954f1a6a17af0db5ec9149e,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,152.15,152.15,exact,canonical-source,profile_8fb9ab5df9eef640038582d9,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
meta/muse-spark,Muse Spark,Meta,closed,52,arena-text,Arena Text,preference,548fcd7a18be4345693ebf86d576de2d43019f937ac45b7a4f82a1a121df6263,363fd42ea9d29c7c11dc17dff4aac9c05901bdeabc54bd5b74e75db28cfeffcc,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1489.0250787876926,1489,exact,canonical-source,profile_54436950921b18ec0c2e019b,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
meta/muse-spark,Muse Spark,Meta,closed,52,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,fca5efdd6a17266095df90ba2b32c6e5be2c8da4c5cc74bfadc738475ece3bf3,371f5f8915bfa5c5af4ec346e616ba4e279ec83663e676e52b4c676e5c3b2426,9234af8f785e13a8089bddeb8d5d199fd74451763ba1563374e16d789e678707,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,40.56,40.56,exact,canonical-source,profile_353e71505b6265ea76a3923f,reported,unknown,,unknown,,false,2026-04-08T16:57:23Z,stale,206,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
meta/muse-spark,Muse Spark,Meta,closed,52,gpqa-diamond,GPQA Diamond,science,a807c4623b3e8b45f92db984e4cad279029cd153e8ca5398b02e93ffe2bccabe,c8cd5336606e9672becb7fee4b80de6a7e63a0d17e40fe2b47a4e36d303e9f28,8eb94ce1c7eca4bb5ccb6ed3fc544b3506d0f84ddc2e21b3ba13374795c6a4d5,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,89.646,89.65,exact,canonical-source,profile_7503d5cd94f11a21bb05a963,reported,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
meta/muse-spark,Muse Spark,Meta,closed,52,mmlu-pro,MMLU-Pro,knowledge,1c6577c631648ca00b15a89b33ba72d6a8d53e491001e794079ce61607663881,c169d5759e81e0a5402ad68162cb8d8e552a8a0b470826aa047f5f637c440902,488b985a7c1e9ad31086aae5ef1dac2f5c490af43542568bfbf68d78e6c74855,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,87.316,87.3,exact,canonical-source,profile_1f4805d656bfdb6334fd4c90,reported,unknown,,unknown,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
meta/muse-spark,Muse Spark,Meta,closed,52,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,494277d066064ccdf37f29ce588d0f4cb39c6faebe40430d65fd9d174b0e5164,e6d038725a8e71601f5b1bd8741ab12a13c8096a71bc9505ff5b2520d28208bd,828ff4b76440946102b6be1ecff774931e036c25110ae2a236ffab6b1605b468,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,88.9,88.9,exact,canonical-source,profile_34bc8136627a1b5e0353c16c,reported,unknown,,unknown,,false,2026-04-08T16:30:04.184Z,stale,206,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
meta/muse-spark,Muse Spark,Meta,closed,52,mmmu-pro,MMMU-Pro,multimodal,35a5006f32850ff4cfa34e4351ad66d7c52d63aca10e9d052b90da6a2b284048,f76486391beee52fb05848249a61486ce555f5ee703c86a630e8042b4f5e1820,75f6dbdd9b3b5a4be5d3e13162b6b7beae758793ff8e63f79bdcd40dc99a4d0f,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,87.399,87.4,exact,canonical-source,profile_8166611feaa70b222ae62c85,reported,unknown,,unknown,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
meta/muse-spark,Muse Spark,Meta,closed,52,critpt,CritPt,science,13cbf17052cddddc870d5bb604db6704740b717caabfceb515a8ba4492bfcebb,0784446646e931b2df38416335ce3dcaaff2d4b33e89076d1c1c7f11d459b584,65600d09ac84b2d20072d49a794cfb11bf60e2bbae25e4f9cc8ef0e82c365666,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,11.330612244898001,11.3,exact,canonical-source,profile_5638767fc11cbaf7aa5af326,reported,unknown,,unknown,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
meta/muse-spark,Muse Spark,Meta,closed,52,swe-bench-verified,SWE-bench Verified,coding-agent,8447496a87e86403ff6a41c23746c738916928da6da0eac54b78d8af08e643e7,3568c676ff6d77664f19f8e27854ec5c1ad8b81c50df63fbb3869682b24f3b76,7d69ea4737f37bcd9bc25b6bfac9c4467657b2e6f16c887fc9bdfaaca48c26fa,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,74.4,74.4,exact,canonical-source,profile_c2ccf01d838a7e46f5423db1,reported,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
meta/muse-spark,Muse Spark,Meta,closed,52,scicode,SciCode,coding,e8a5198327ec04b34163dfcc54655701b1b35fb0e5f90532432b87cdaaa9ac90,c90f43397baa0b60f64d13aad84c3ce977ceced0f8866b305d859bfd2083dcd4,ff35bb91ca728f3e1a6604a6494745a4224ae935d13a0e1e1741a70135b63771,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,51.5046296296296,51.5,exact,canonical-source,profile_7c87f12cd9a3ae0363f7226b,reported,unknown,,unknown,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
meta/muse-spark,Muse Spark,Meta,closed,52,vibe-code-bench,Vibe Code Bench,coding-agent,7a6bafb9fd8be57ed41b4ae5fb4764a656cbe8c9dfcf778e9cba59664eea47ab,8bd1d25523be36454ad0272dfc06a7d3de4f35e28dd20f5ecd1b8eb92673dfee,f2a6d59e558bf9711e27caff7908e4b9b45d55c7b9a25d02f9f628dd62ad98db,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,19.674,19.7,exact,canonical-source,profile_caf27ae6f0e02b627cb46007,reported,unknown,,unknown,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
meta/muse-spark,Muse Spark,Meta,closed,52,webdev-arena,Text Arena · Coding,coding,90d65fd5768874c5bd7a1520c4e48ada0fe66e7a7c619ce23b0bd5ac19f5a643,7dd26befd75b7f0fcacbd2120f81958cb4133fb70977cb9871b18aef48869d80,7478f140435c0398c7174a5112f26ce28b9af101e641529581b4f498d38610fe,release_65fc140e7ffbb7bf62021f84,Text Arena Coding current pool,text-arena-coding-bradley-terry,bradley_terry_rating,points,true,1512.53,1512.53,exact,canonical-source,profile_3bf852297bfa61c449d25840,reported,unknown,,unknown,,false,2026-08-28T22:47:56Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-webdev-arena,Epoch AI mirror · Text Arena (Coding),https://epoch.ai/benchmarks/webdev-arena
meta/muse-spark,Muse Spark,Meta,closed,52,mcp-atlas,MCP-Atlas · Scale,tool-use,48b125771298d76fb563182f486d8d07f65f7c5c789695b829b5c9b876b39cd2,780283dc428d6c54e242ed0d4f8b6fe6e57fc4a6fb8f0c4a498849b165eab06c,dd19bfccf6bf1c77eabc9c426dcfb2b3a0ae56ab277763bd692454e86a76a8e0,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,82.2,82.2,exact,canonical-source,profile_c20d16cecbbd58186fa001f5,reported,unknown,,unknown,,false,2026-04-08T16:48:43Z,stale,206,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
meta/muse-spark,Muse Spark,Meta,closed,52,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,484ef473b7c75e4b6e708b36492684ab3d75c9c2a921df456110fd28479280f2,22bd6716c75805b9e2aaac5e6d510b3d9489e811cfd18fa8d355ed3635d5421d,3514805a04926af1a9e75d945bbe905d97eed2400f8ee962bdae38ac8c054ff1,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1006.05,1006,exact,canonical-source,profile_d20f8a5e90eacbc5f8ebdd68,reported,unknown,,unknown,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
zai/glm-5-2,GLM 5.2,Z.ai,open,53,epoch-eci,Epoch Capabilities Index,capability-index,b73598db6a56a4893df2e28e04b147537c04faafe711f4dcff1dda47b60b3340,31de7d52afa6ec6856ae841bf9b64fe7beda9417f4de6a27605cc1e1bd1e2244,b3816f11696cb8e3e7ba878f4e89cf8ad004083419a05de1b3e82c71bc4f6ca4,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,152.07,152.07,exact,canonical-source,profile_04566da2338e5d8f2694eef1,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
zai/glm-5-2,GLM 5.2,Z.ai,open,53,arena-text,Arena Text,preference,c77ffde998cae4f5dc55e016507ad16fc41cff23652ee9bb4c219afd306741fe,1cbb8bb4cd74f49c4123cfbc12b00ffc2e700ade20fb3dd5246ebfec3a90716b,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1475.289755722248,1475,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
zai/glm-5-2,GLM 5.2,Z.ai,open,53,livebench,LiveBench,general,57b7e7b67593dd884f3cbcae7708305e3d7fdf3003ab58e1726b8f440f31dcab,25b911f9e4ec247bb1dc6b80dffbf69a02c3321a6f8295dbe1a46b4be380e87b,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,73.15728571428572,73.16,exact,canonical-source,profile_17119771e437727ff7cadce7,reported,unknown,,unknown,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
zai/glm-5-2,GLM 5.2,Z.ai,open,53,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,f6dfb4107995dc93dc2f4d4b88fa14e3b55170809fe5be0345a9fc132ba4bcb1,16f3e75e23e33f4fd48152976a0f47b27678bbab45c6786bd0495477854d8971,e1f956386a97e6ba42a3c396757902d19f19cdabb10fcc91e345af5c771a2890,release_fa9e908ac04b067eb00182d4,HLE official leaderboard,published-leaderboard,accuracy,percent,true,40.5,40.5,other-supplemental,supplemental-source,profile_c54b095bc209fa08007ec46c,reported,unknown,,unknown,,false,2026-08-29T23:52:43Z,fresh,63,1,A,1,A,5,C,self-reported,6,self-reported,hle,Humanity's Last Exam,https://huggingface.co/api/datasets/cais/hle/leaderboard?limit=200
zai/glm-5-2,GLM 5.2,Z.ai,open,53,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,b9dc0bd803e415cb3dc0627b202af3c2bf6845184f05648f73d4a9e56b98bea4,f3c57afdc7d1b2a7aacb0be4cf36e8d39f6caa48d960c2e42e64384a4b2c0845,92ed48917c513acfb6e6762c693dc06dddf60cab9a314f607c515ab653f67106,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,41.149212,41.1,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
zai/glm-5-2,GLM 5.2,Z.ai,open,53,simpleqa-verified,SimpleQA Verified,factuality,887f6314b80e68300f8881661cf8f4e9f21c73f35b156979376cf61d3604ba15,0d63f637af49da91f99d4b7ff69dc59d1d360d7f00e77e6a4fc3ed8c474cbe45,793a7e19bb4532d6fd827caa77a0e7c75b6fe3c7661e7cb6dc360e76b8ab1b4f,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,34.2,34.2,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-08-27T19:36:05.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
zai/glm-5-2,GLM 5.2,Z.ai,open,53,gpqa-diamond,GPQA Diamond,science,512ef10152bdb77490b44f0b5ab6581bdcefaaa401401f6f950c8ed9aee46efb,e2ea70f54b451f5ee8b58304ae35efcc2c600f1f093df7b11725e8f998f40093,e78e34ddf169b110243367a84e6552e35050e46eb492a50a41a03147cc0abba4,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,85.606,85.61,exact,canonical-source,profile_43b1a4c9d140c7bbdd7604dc,reported,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
zai/glm-5-2,GLM 5.2,Z.ai,open,53,mmlu-pro,MMLU-Pro,knowledge,cd65da2c458fdd5736e069aa85775f87dd80430acb157df52fbd506231673d85,0b322b988f4e472d23d6eff59f145761c15e1b9ffdc646ffc092b06489a1ef8c,85d72554edfef33ae0f6c22cbb1638bc9e295243b01dbacc99a718cd0a9eda9a,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.714,86.7,exact,canonical-source,profile_21a68f413995c1824054e5f0,reported,unknown,,unknown,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
zai/glm-5-2,GLM 5.2,Z.ai,open,53,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,f076089806a686e0695391ea6b2b5bd3c3920d9d4cb0fa0ecf4f75a18ae86602,dac846086da0be0a58d5ccd83d413edc80316873f309ba09363449126bc634d6,15d072ac1920c685176ffc0de7e75f477ea07dbe82bf73801bbd24efbc815915,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,86.38888888888889,86.4,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-06-25T00:15:08.000Z,fresh,129,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
zai/glm-5-2,GLM 5.2,Z.ai,open,53,frontiermath,FrontierMath T1–3,mathematics,ae26aef06935c7563507324dbdda6de55d9aea04a786378dc7743b72f2f837cd,93941b7064c8d503bcb71ead41fc56f7881549c6e7664ad4ca16227b7f417121,2b719d2dc1d51447164fab6e154bf5117436d13b91c8542fa3543c966796cda1,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,59.205776173285194,59.2,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-06-19T16:13:05.000Z,fresh,134,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
zai/glm-5-2,GLM 5.2,Z.ai,open,53,arc-agi-2,ARC-AGI-2,abstraction,e3992e98e916cf76698d29d0e9edd822c3a9dab8904c55ba577d4a29960014d4,bf8f587b3b34ee9a08654e4993dd2cd134683882f0683591c58b3534f3c7c4b7,ee7689cab2ecdae3f0d03646646ae70a43a48a3d5e8191ac81a5dfd541badceb,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,22.775000000000002,23,exact,canonical-source,profile_35d0fa8752189fb3eac4cea6,reported,unknown,,unknown,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
zai/glm-5-2,GLM 5.2,Z.ai,open,53,frontiermath-tier-4,FrontierMath T4,mathematics,c8d8a201c44659ecb20f465ff765b63f1d2c85ca22a15e6cb4e132abee97e363,c5521725ac630b02cf3890dbd975f1585a8b2d6c8a4a2b808510b7324d969b33,908684107da51a0d817af2b4d3f9faf3d656f68c7600d9fae11ffff479060d82,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,29.268292682926827,29.3,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-06-19T16:13:05.000Z,fresh,134,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
zai/glm-5-2,GLM 5.2,Z.ai,open,53,critpt,CritPt,science,288743e29cea37a7fff38008aee987f8f817cbd4c09b158f976e3d523e50d283,3e614d8861fa56bbbf08bad4157f67b080177e8bcff97b91e3c0af57200339c6,74f6788998c46fa98e6574d81e5167a10d43e1335629992c67c2f44169f45533,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,20.8571428571429,20.9,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
zai/glm-5-2,GLM 5.2,Z.ai,open,53,simplebench,SimpleBench,reasoning,6be0d535110e7a5f1317791dbe0a261d117b3786ca5f34c53c8bd3290207d0c1,d15ac7579e3fc64ac60f8920505d3c68e744a2ee7b70a0d527d009228a83978d,fbbd9e0d0e3dae62501f10feb8ae007290e95726caad1c8211fbd34936a787a9,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,58.8,58.8,exact,canonical-source,profile_c362d9b546b44a292da204a3,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
zai/glm-5-2,GLM 5.2,Z.ai,open,53,chess-puzzles,Chess Puzzles,games,750f872f72312e5ecacfd7b4fc8fda8e4a7420db333d7b7f99bbf33bea3c6e59,6130247474bf3d3f934e92ff3a522222df499ef0374947d417ed5ebb04745cc7,bf68acd2d1d1dd5a3ab2ab019a17e25dc407ee5827526e24fdf2af98912663d4,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,22.0,22.0,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-10-14T05:10:27.000Z,fresh,17,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
zai/glm-5-2,GLM 5.2,Z.ai,open,53,arc-agi-1,ARC-AGI-1,abstraction,f3817f812f44b714cae9cd96af933fc7705348b8fe289c3bcd629b3dde47e57f,897b95370c2fd09f6264509d0432cf7d607a3c711315f55894fd5b22b13eaec7,0bc797d563c9920ad5117d503a938d125f8b22dab8024cd5dcc36cb2971a2a71,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,77.0,77.0,exact,canonical-source,profile_fbde99d386073827dc60739c,reported,unknown,,unknown,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
zai/glm-5-2,GLM 5.2,Z.ai,open,53,swe-bench-verified,SWE-bench Verified,coding-agent,699181b7257503d107bd7b262a960640ea5634bcc4d75e7bad0b275d4cac1082,4ed0b4565f976acf81eb06526c3c209db17bdc0ee0bbb2aef560e013e565fd8f,bf256637ab7c031c6ce221713d534fdbcab5facf122e6dc16b3a624a75293036,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,82.8,82.8,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-06-17,fresh,137,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
zai/glm-5-2,GLM 5.2,Z.ai,open,53,swe-rebench-v2,SWE-rebench v2,coding-agent,a0a57cf3d28bcd91bd306d1abf514f7107f0486cd56a6c7d30c085709d1eb8b3,2c185be4b4fa421c03cd92083933a8ef1b134e9238eac4d25873241b33bf5bdf,ef3c892e6111a7703e93a922e80e35358681f631e166e5fcacaba61b8a3f3d38,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,62.88288288288288,62.88,contamination-warning,contamination-risk,profile_0a11b05baf509af0577d53c4,high,high,true,fixed_effort,,false,2026-07-01T00:00:00Z,fresh,123,0,A+,0,A+,7,C,potential-contamination,2,potential-contamination,swe-rebench,SWE-rebench,https://swe-rebench.com/
zai/glm-5-2,GLM 5.2,Z.ai,open,53,deepswe,DeepSWE,coding-agent,e07424058e54e99848dbd34aa3149708cc2ddec67c0f70687d75563f39da3deb,53857893dabf0f8bdf12a0aa54e6e5ff0538ee519f53155c070c28c8759178e1,7c64f1205ab13b5ff20787d49fd0c8b5dda15db5f0285f7bad596dd073bcb96a,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,43.77777777777778,43.8,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
zai/glm-5-2,GLM 5.2,Z.ai,open,53,frontiercode,FrontierCode 1.1 · Main,coding-agent,f0d9d3c0b954e5e0e2706c5e9d94db8679127c3088856cc80f4d50c271b5c6cb,19856017c6b6ee034cc9ceb370eacbe0b6f1beb5acf09df81796b0d573cb514d,6c9a24d40701bbc2cd3467fabf75eb78af4cf0a4f8ef7cbc7fd8ef9023d0dd90,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,24.5,24.5,exact,canonical-source,profile_2dc4d509d97ba2d565e36ba8,off,off,false,none,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
zai/glm-5-2,GLM 5.2,Z.ai,open,53,livecodebench,LiveCodeBench,coding,b06976552a60b413b091f62058113dbf796d71d5d24666585cb50b55aac32496,1f0b0ce1708028c09b05557f856b28d04f0e7fc4d447c8defaaa08d3782139bd,6890c60f64018dba271634c7a4f7252db6a28528826624dcef0d31028896de19,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,69.5,69.5,exact,canonical-source,profile_7fa74e2701e012fb5d803f8c,reported,unknown,,unknown,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
zai/glm-5-2,GLM 5.2,Z.ai,open,53,scicode,SciCode,coding,48fad2d3268780bf5771aff570253e7996fe8d9203eb7f571d8336e25f61fa0f,0b1ad08946d0ac857c8d2b8266ac045fd42ef5f139384aac7afb13159e169009,9c5ff701c4d182724e8ff1d1149e16824f94497bc929364f5ab62d237a85eb48,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,51.1574074074074,51.2,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
zai/glm-5-2,GLM 5.2,Z.ai,open,53,ale-bench,ALE-Bench,coding,5b66724705786655a4bb8a26d3e0bf542525d88bd00721d74c0af1f118ff0e70,b8a27406f4199a328f7bcbb5c6fed5ba4c93ee3a90bd5a32c6cbbd847a4d6c6d,4c1cf9e791c0b26576c00a51c42d25ce96855e5afbca22f2f22be725da0ad7e0,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1010.17,1010.17,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
zai/glm-5-2,GLM 5.2,Z.ai,open,53,vibe-code-bench,Vibe Code Bench,coding-agent,5ef4abb4c5405144cf77aa59c296ad23ba4676245bc49881748c83379be92234,777284d28cd2eeeeaf88b0546a38c17f9a57a63526e75154c460b30ebb7ad1cc,90d16c84acfc61cfe588d342864108a487f6d96c8dc17c2e54d0aee03aa20b62,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,63.961,64.0,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-06-17,fresh,137,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
zai/glm-5-2,GLM 5.2,Z.ai,open,53,webdev-arena,Text Arena · Coding,coding,b67ec8809db01575f08d5ddacefb2152394deba470936bbfaeece1fe547c4c52,466ae71b334ded0e94eb668f5e836c63ac20ffc308c7ce6908206b9d060a4677,554ee0b55045d03ae4f0841b21ce34e16bec34e23700bc5e40b71d2decede08e,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1605.639209438289,1606,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
zai/glm-5-2,GLM 5.2,Z.ai,open,53,programbench,ProgramBench,coding-agent,d0111b10856c844d05dee5aba9a3990aa24d04f4fd4e308b7f01eeb0575f7bd2,9821a9721465c677c4b3a1dc214cd15b04397fdd2df91b4224105f8857567f23,0498d1d4ec0b44e63a1b799dd92f4609f582b4309c3b6d7f1e5b6b5c7b58a6ee,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.5,0.5,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
zai/glm-5-2,GLM 5.2,Z.ai,open,53,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,5df6aba2d10250fb2d0ce62f2506a50d70335f9b33685e7e4143421d59b56610,3152fc91790200d542342d18ebee111beb6f1002beea5f91e83f4b4ae7754d9f,9aa3cc7610c1afcb4b30fa74e682ff351eebe92aeb2dc21c15821014d832ba19,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,67.79,67.8,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-06-17,fresh,137,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
zai/glm-5-2,GLM 5.2,Z.ai,open,53,sre-bench,SRE Bench,terminal-agent,e78384cb1fe0872fbf52a8a87ec0bd472b094d2c30e72763ef2c25ab7565e503,5aa007d9a7efecb7c86a0201dc15602d19d6c1f6fe7613991ba91add218a9f0c,d11c1b5a93479123d6856b3b4e086745ad082b0349b8837d9f2cc5091096620e,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
zai/glm-5-2,GLM 5.2,Z.ai,open,53,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,f7f37d1c8f98033ccd800768d437d53aa57aa2c934a44d1a9c3fa390c4e72a56,945012d71431e412d45ec6b0a12aa17eb6e71c3e02b18d5f4834f1ef403b6a3e,4c59c099162c4a01f2649c2a1a8acc544f302d9a3204752aadb34fb996add352,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,34.639175,34.6,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
zai/glm-5-2,GLM 5.2,Z.ai,open,53,mcp-atlas,MCP-Atlas · Scale,tool-use,8dc363bf17e56ec5aa41c3abbca1f3ad0235bbd2b20cb51f31817a832b6a26f1,98af08760f6f81aecfd0425d566885646f6c561b75633ca9729b8b5a86c83b90,a4a445ae106541276befbbe5201f6e9f95a8556aca9b8a2d68f553612f5e495d,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,77.8,77.8,exact,canonical-source,profile_485b19c39bb21348a203eb53,reported,unknown,,unknown,,false,2026-07-20T17:09:57Z,fresh,103,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
zai/glm-5-2,GLM 5.2,Z.ai,open,53,apex-agents,APEX-Agents,tool-use,0b629c408b95276a12a2fb9308b636d137b3913a4fc5b4603fe7e00fd1e65443,57ed3b868adc7190962560cea289fe4303feff166fda63f3d02f83cd2b9df7dc,e825f781a78eb80c3ea82eecae389e14eeb60209868b63e1f166b59662c1e788,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,45.2,45.2,exact,canonical-source,profile_27cd4b5c9a4a27c645ccfbf4,reported,unknown,,unknown,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
zai/glm-5-2,GLM 5.2,Z.ai,open,53,arena-agent,LMArena Agent Arena,multi-turn-agent,ab5ae7b983caa4a8af9424c3d85a51a849c08a48b6f29454421574f00483731d,e1494bd3e9ab47acd11ed9d63a28c265fccc585c3ffc18e778d44ddf6975317c,1cb3309b6d5b2110a73414c995aa350cfe90961842152694d60ae49890cbb3e9,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.012550529818184046,0.0126,exact,source-specific,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
zai/glm-5-2,GLM 5.2,Z.ai,open,53,vending-bench-2,Vending-Bench 2,multi-turn-agent,50b351c689b98ae27e7211c8bde7b868fe55135ff00328fa80c87ec7db221f05,013cea20f1a660a1a55f9d59c72300b49efced23146fbc93e315a10abec58592,30bb57dda941478e098ec5e225aa9517aeeff59a03ace8296132a1a40706b998,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,8313.778333333335,8313.78,exact,canonical-source,profile_20ab07deb4f86981c9497a7c,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
zai/glm-5-2,GLM 5.2,Z.ai,open,53,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,f62257abf07a529e0d4c4e150ce87f9eeadfbe3a4e5445714e5537e68f6b3a1f,89df3082d21e0769fa7016f31bb508ec0c5fa02e2f3f9550ac511be54e277a23,df51cae64d9007c26a3fb899335384920dc40dfed17fc094fc27bcd22f51de16,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1381.49,1381,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
zai/glm-5-2,GLM 5.2,Z.ai,open,53,terminal-bench-3,Terminal-Bench 3.0,terminal-agent,f25aef2b4ae365ee4e57dc7b13e846600b54cab946f4165a35d2b864a2b8e986,6d26e312adf3154eae3de6b419df7d390cc6f7fa31e2c86f95cfac4740fa7613,8193d21ad0eb789a1ec26ea50fbd3379bed4aad4e8b492bd1b886fd1b07a3897,release_ceb36f25882d8cea2c9d084a,3.0 · b936386b-4afd-4ad7-aa7f-6b52bcab9853,official-74-task-leaderboard,accuracy,percent,true,4.59,4.59,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-08-12T21:39:34.661007+00:00,fresh,80,0,A+,0,A+,0,A+,official-leaderboard,4,official-leaderboard,terminal-bench-3,Terminal-Bench 3.0,https://ofhuhcpkvzjlejydnvyd.supabase.co/functions/v1/leaderboard-read
zai/glm-5-2,GLM 5.2,Z.ai,open,53,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,e76aad992786b59da7894b4ff8e53b9c3f1a3ae58a28ec354f91d05480729a79,96141a69e70b1b53b101757f601cada71ad6f50f91c8fed6ccca7f4a9ab9954b,8ae31a68885f85099acfd11744c8b6482786a6c228b5cf38146830c9fac67e2e,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,37.11340206185567,37.1,exact,canonical-source,profile_731f1098dce287a031491f2b,xhigh,xhigh,true,fixed_effort,,false,2026-07-24,fresh,100,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
zai/glm-5-2,GLM 5.2,Z.ai,open,53,weirdml,WeirdML,coding,224309fa504415cab2525617b28001cdef1acf4c105677cb084d878e6c23cf50,9dd1f5060edd4a97413e923f608fe28f59cbc22e16c41723e3f625c7cbef321e,8c398aaaeb7298f9c247a193edd1cb9b66a708611a8f3497fa2c2efe0634120c,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,70.12,70.1,exact,canonical-source,profile_a6e3ca6e600d7d7ad3656281,max,max,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,epoch-eci,Epoch Capabilities Index,capability-index,a59e52a24ee576c12d5bde36ec917023aacdaaf45e277110dd7517657139fa11,bb58a9eba772ba51270cd3e3eb16eb7f0c2e781d5c43d32490f6465a6270259a,d36ebd95915c2a8ad6edce1dced434dd6e2ebde185c64651559ebe84632f0954,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,152.01,152.01,exact,canonical-source,profile_9ffb97cc9acccbbc83ffb92e,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,arena-text,Arena Text,preference,2df345827c77214c2ca7237a728413b4b5a2d0ba060c935df985c184954ccf94,d5bd0306f3d39c23745a76e621b4e98e3f0e7664a8a45b30c78032f23f1e0966,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1458.367316107511,1458,exact,canonical-source,profile_57d19654e2ae916ecd0b5c09,minimal,minimal,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,1328f88b0775ea781b01ca80ddfee244a6e2e349952204872c2029a0deb25c06,3363893cf9e742cee93dda00777449346362ce7e2e56df2d111008ceacb3734e,26c8fd00871cbe3c4d547eae32d78e6ffec3b09798c45289975fdd0b75864645,release_5f3023448c2ae4632fe1a43b,reported-runs,public-multimodal-mixed-run-configs,accuracy,percent,true,36.6,36.6,other-supplemental,supplemental-source,profile_a37ea0078b8bcb4cc5bfbaf0,reported,unknown,,unknown,,false,2025-12-17,stale,319,2,B,2,B,1,A,evals-report-official,0,evals-report-official,evals-report-hle,evals.report · Humanity's Last Exam,https://evals.report/benchmarks/humanitys-last-exam?tab=scores
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,simpleqa-verified,SimpleQA Verified,factuality,03c5a256a2f869328d699943e908205834fed8b55194dd209c8eea65fad033f1,bb93b3e4e4fc7f673cd6ca989d7cb87fb34ad21964a264f02dbe67e769f5200e,05f29e9b0116a60ecf7ad7e2990a0fdde7e3db4fd096daa36075437027847a71,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,66.8,66.8,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-08-27T19:30:06.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,gpqa-diamond,GPQA Diamond,science,542df8aaf17127c731fdbaf75b457530c404cfb66728bb5a62f2b21fe14a7568,3e7e850d5233073abd7bcb706d6b2f3fc8aac4a16b1653a580e0200b09582e39,737c99f1149726c71406ee9ccd875de1bbb060addb1bf113a504ba2a11f16b99,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,87.879,87.88,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,mmlu-pro,MMLU-Pro,knowledge,add417da102d1dbc28e763805e89bf632907796249fed14501eeb81f2710a86f,c902955d5440e81f1c1bfbb1a32ec5a21df091f29afd05a16b46c1a8203d675a,740646b0391029bbbb479a7d17da7b6e53ff9aadf90a736b49f3a4ade9288e0c,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,88.592,88.6,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,a5b7732bb496ddf59399e195f4c1863d8f784f3fe9f07138289ac3d3cdfedb12,2a2715bbd4b363851334448e5d992f78eeef6af4e567ad709ba412952c77067c,ee7434d0c3b0d41a21b877f0967b93eddf632c1e85f5c451cedbf79f383c609d,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,95.55555555555556,95.6,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-08-06T22:25:57.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,frontiermath,FrontierMath T1–3,mathematics,483152cac0547daa61a2855db08a442f3d6a2d1fbd36a2190b92456148d9dacd,69c127ea24cb5dce5fdd4d85924891031f8e6bbb80472b57a123c9c1c12c2bda,96cd26da29a9ae656aab7d662b37cde34e048e9768f2ea03150eb5a79fcd803a,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,51.2280701754386,51.2,exact,canonical-source,profile_0ba24c793659a66d8cbdc90d,reported,unknown,,unknown,,false,2026-06-11T23:44:30.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,arc-agi-2,ARC-AGI-2,abstraction,7a0b468673edf91acf510b8af10039e5a1b59b266e97b13d1a3fcb7ec581a25c,4ea646b0e9cc3dec38df6b3e8487ad885ce2f864af5f85e3173d16d5f2f24eae,f8312f452c1590fb2d8d5d0058ebce8a3f1cfed597e44f9ed6968b16bd4b9dfa,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,33.61,34,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,frontiermath-tier-4,FrontierMath T4,mathematics,6cdd70fb231105a87661ac862b6485bbd3c03a3081d751e4deab4d1a705b9220,a30f13062dc47c611faf924eb2176b1335d1a7e6f83791331764d89a025f4608,0ff253f7b5fd2063e6cb521ab31fc665724c492b9c6f1484984fbe7bddeef367,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,17.073170731707318,17.1,exact,canonical-source,profile_92976465691e2c940c861d92,reported,unknown,,unknown,,false,2026-06-11T23:44:30.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,mmmu-pro,MMMU-Pro,multimodal,8973f1017aacb97397d8a74b475a4b20abd1f478508ebdd1bd254e4d7d78383f,136b6516992aacf16747db97ae2eeede950f8a806a50f3062199ceecb19bba9f,7882730b62ce67227923945d91d9b57b6413cc45e1dea45311898ada3ca0eb8d,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,87.63,87.6,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,simplebench,SimpleBench,reasoning,69b87bffea7346d93188749ce43bb8c7c1ad8190236c9cdbdf5be8726f5e4e2a,bed09e5286fc298fda73e7e601c2adf807f6b7f66363f888b0a8e9932f0e8430,3314247dc49fdff2278501e4ff87391a4638f65db8f3227c6256e5e39a236aae,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,61.1,61.1,exact,canonical-source,profile_d6963c4fa76fa587ee8fc818,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,chess-puzzles,Chess Puzzles,games,a90f7327b5e321cb56f848e0c34af1b14316e5d90f3e2bfa46061603478f6f68,dcfbadf0ca69e0c505bc7f2c5327a0614a158319b53ae1c7d1aa44e15df8e437,11a80a7acb36a3b00a89f6270c751bd02267a824b332a75adfe71aea5d5102e3,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,39.0,39.0,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-10-13T13:42:10.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,arc-agi-1,ARC-AGI-1,abstraction,88a2cf2fc17bb2636acda8e51cbcebf3d1074fd32c94344aa6c91536946cbb01,c38458179c8a70105aeaed946bac67e97018a9e3854846eb61d685ad972ac19e,c5ce19de294519af53926129ed4992f7ed8e37de295da6aa5c8ca2f057ba9cb9,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,84.67,84.7,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,swe-bench-verified,SWE-bench Verified,coding-agent,27fbf1c8baec83d373c0ce0f3d12091022506a5b004f627bc5dbe5c1081e7053,e36d5439ec9c4cb9dd250b28ff12b712449341ff7d054d40460ddebe770ce154,33ac1d31a772d74c58b6e9ad2d6f823170018890c79c206152a471b6b184a331,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,75.0,75.0,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,swe-rebench-v2,SWE-rebench v2,coding-agent,ffe691496fde7034c7b5b18ba3e6a219fcbc81444b4106d2378c9f613dcc8ecb,f995ba60151bf9bbe00b9e42b17e2ca9dcf14edad9220040f25a4868f5e9f279,41b89107cff12d693d8637505f026e8beeab77a4802fcd952a837ac613bbdda4,release_b31e2006211423b08d16b59d,rolling-v2,2026-01-01/2026-03-01,resolved_rate,percent,true,49.81135531135531,49.81,other-supplemental,supplemental-prior-window,profile_e83e7303a3376faa702de988,reported,unknown,,unknown,,false,2026-03-01T00:00:00Z,stale,245,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,gso,GSO,coding-agent,1ce28c8c910ab4e3b3192ac1400da75d3c7308a663e8c44363798b1c44f583d0,7cb57f5613fc098aef1190cbca4e117c8cc54f0dd5e8498aefda0b8ac47d2849,c20751fb7a44d040c598104ce00cc785040a219233ed3f47b0b005fc63dff5d0,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,9.8,9.8,exact,canonical-source,profile_525bef63a282c86f51e299ca,reported,unknown,,unknown,,false,2025-12-19,stale,317,2,B,2,B,3,B,official-secondary-verification,0,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,livecodebench,LiveCodeBench,coding,7fad1e2c98d50b21ae1c22c9d0c3c5865bb5197e23ee3460dd8a6a2c02c0bbb8,0ced2e793ef9f91674865df20e70511112ff9c7805ae9d4a9ab053355330e5c9,36fedcc4dd8a8e2034baa70273cdcb97304308390dbb24653a1553ce66b20a20,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,85.591,85.6,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,ale-bench,ALE-Bench,coding,0b5224b2da2aa5baab1bfe106126ab300a363e7f0e7bb58dae8cc80c814d9f70,253ef954cdd6438bf3af944b6feb56bb200430e8f9b2be780571712893d32976,3b90e109f668919c06399d02a95b3e09bad004e7366346c6cc39d5c44dc328a8,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1367.2,1367.20,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,vibe-code-bench,Vibe Code Bench,coding-agent,1bdc159788a93f42126e872023031baca4f279afc0a6945c2e40bb2c2249f846,9b16e0ab06e83c6b03116c8cb83baa875635776bb332d69ba4969d447bb409a8,96778b7801a377a5e232baa65e0b0884c17061f9bdc0b804c4b364e17647d406,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,20.204,20.2,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,webdev-arena,Text Arena · Coding,coding,a15b359f28b37b893d8c5cc81afdfad8d66b5ffb9ebf0161a3f506dbabc58db6,075ba180e933d115d65a9e5581714e323426185c6ab40f754fa258928dd89c76,e4789560e0f2725c957333104d2d603507fbc04e3b8942226a4185824881a7ff,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1384.4938378264999,1384,exact,canonical-source,profile_57d19654e2ae916ecd0b5c09,minimal,minimal,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,programbench,ProgramBench,coding-agent,afa809aaeabb0144cfa90a73c7ecaa2e514d37384084730df438c54245cd8b63,7c74cde97fad43c22428930339c711ad9e7bf1c1968ee48618aae907a8baa8a2,7cb216d52eb88a3debefb4d02a2f246dae971db4823c6e0528bc1843f043ae35,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,5bd622d09a0718853182f539798270083fa665e45595fdf8c4a05c4e482600a6,62ed62cb3c934097e16af4749897f65f04753b0298b40cec891bc5b310ec2caa,def158633bb02d1fcbf46ae225455ec309684f39357003ed5d57fe3abafa472f,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,53.933,53.9,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,skillsbench,SkillsBench v1.1,tool-use,e1f9e64009a75c086993ae358540a7901f1e156b108aa23731e624e6f7d05d71,f1096254a2e3b14483283fe2787066724e728caa86ac256ba3707c65b05d0b73,b364c19f09c3235a2849dcf46f5f7497a801c57e4a3db1ecf5efd428de0c8418,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,54.6068,54.6,exact,canonical-source,profile_24a77b9d93725e789667e3fd,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,vending-bench-2,Vending-Bench 2,multi-turn-agent,b8945a0fd64232dc7996a42fac2a2316aff85b513899c1254c872a44922f18e3,ebe63c4454fd54da024d361b056c383dc23a6d320d89e1019e187cbe11f31088,b62090d365143d132738ff4217234b598cb3a10704b5d2fbccbb260b57b3516b,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,3634.715999999999,3634.72,exact,canonical-source,profile_bc2d3b57176164ba1eac2f22,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,124c9e2cbe5409801996892616788d6ea62df535c3204a655de06ae9c5cffb70,bef732cb790fc79463fdc4cb595b6d1e7175d4be85575ae76f93522c8329f6f5,65cc8cfab0984dda436d298ef1a18fd2db48c8b2d5c686cddf0f37bf4300de51,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,27.32,27.3,exact,canonical-source,profile_385ff75e5783055d4ca2f505,high,high,true,fixed_effort,,false,2026-03-02,stale,244,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
google/gemini-3-flash,Gemini 3 Flash,Google,closed,54,weirdml,WeirdML,coding,f580756accaa82461b980c8acb1d9b1fc4b4f557ea2e1613604683758d62aca6,9e5a8d9462a3c451cf3260c731111b922aa01bcde26918731554589738aabdec,c01698a40f8b0cbf99e6adf29d614a2f39760fcf6be2ba1feb28f0cfd3503b90,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,61.6,61.6,exact,canonical-source,profile_7572b28212257731460cd760,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,epoch-eci,Epoch Capabilities Index,capability-index,874c0e93f023da75a0913ca483dcff68cefeea0c70d4a8f04c35f7d02c1a1139,3b2aa86b7557b52b16050935b24b86dba8c32e205a1e9bd9f0c224fa94091b0d,a2b66585d4fb313c0a42f3df71cdb16702007590c303c6b03b21c953655145d9,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,151.8,151.80,exact,canonical-source,profile_e182cbfed410a4de7a2d42dc,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,bbffe5fd00bba0e480fa0e4b38eb5db3fdff35048142ca4093cd72813b55d25f,bf78e68170288cee66caec6c4cf92404bcf7a81fd62806af15212f1fd86cee82,e5ff73a88b689a809909cb15a8327c9d57ff18fa7acffb20a1f9e9ec7c674b3e,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,20.106876,20.1,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,arena-text,Arena Text,preference,e1f9067ebf77758ca348f5307ea3babdce35a306167acff1f690ef5597c8e859,e17018a162bce7cfe550e1345dddcc12f616681d854fbe41e0df6627616769cc,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1435.5994372146474,1436,exact,canonical-source,profile_c384460647f97ae85afeb51c,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,livebench,LiveBench,general,c56677c8e3dde196226e7093139304d601864863bdab0dbf6703e4d4a3178b55,42801f4217e7367fd6b36a9567f94cc358af5da961ef7f2875886dc8972476f8,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,75.26910714285714,75.27,exact,canonical-source,profile_da362f90a33ff4c7924ed94b,reported,unknown,,unknown,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,72d839b6fa6a023ae351cb1baf4c4f51eacaddca4530e13b72c87e577729ad9e,382269a32ca84b946eb107db590b7b7c69df58670099d25d7d45c8bdf4a6fc21,c93a278e8ec3e88d3cf8590bc92e78109edbf3fa8907cc99906d802003896656,release_287aa4bcb31b5b1ba26e562a,HLE · Qwen vendor run,vendor-card-protocol-undisclosed,accuracy,percent,true,30.8,30.8,other-supplemental,supplemental-source,profile_f97c4e36649aacbbeb6c72f4,reported,unknown,,unknown,,false,2026-08-14,fresh,79,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-8-27b-card,Qwen3.8-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.8-27B
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,5a9b232404a29412abfe265903fd8d2f948f747546ecf3eb2796c7fc2405b6a6,2fcb980ed31968835d39347a41d180077ceb43d9e77d08b793651995b379c3f9,9a73eba5f5692f957518717c74be4084fd8f4cf2bc7fcdaebba1a1bfd2c89625,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,33.920297,33.9,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,gpqa-diamond,GPQA Diamond,science,a12a5689fbeec4dfc278bd48eeb83fc0493adba56ec03d5374435feb0d8e989b,abbd50e9a8b769ef951f75782dbfd8ab5d234d2b5beb5da94438565b82e50f36,7d010d340bdba9bbef6871bd34f0dc3c1070a1804b1fbe100135eb17f6597a32,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,88.889,88.89,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,mmlu-pro,MMLU-Pro,knowledge,344746e8221bd455f62b8d9bc9cec8a1dce2a761bd59221d570be9c0e580d9be,c20ead303a9f543a5d2cd66a0f6c5370ebc04887472d0c4a3361e678176a2a20,78050e47b74a82aba0778659ae32272c9acd82d3dab9b593ad89321878825832,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,84.337,84.3,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,arc-agi-2,ARC-AGI-2,abstraction,6ae153f9f44807c7838b68a834c80a76e0675a427594000f1481cdb8c16ceb96,550348c0f3814e6f9765f6e21dcb377d1e212053a3c218b63c29a273eb846d0f,3a51d79cb6ec2f0dad6101a036a1aa9a9031c99698e9e69ec67d4b2b40a089b3,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,42.36111111111111,42,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-01T20:29:26Z,fresh,30,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,mmmu-pro,MMMU-Pro,multimodal,8e000021fbbab64cfb726102fa61961041147495c7433403d9b5e190ce05ab4e,6707090f55c8c4004581d52edca44d8af3bcd65092983aa2931209ea3cec6b21,385c6722c9bada8ddde7aeb4b35c9ed8084af42f46e15fc996b9b0663fc95c2f,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,83.931,83.9,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,critpt,CritPt,science,13f22cfb731021fea43aab01105acd8b50ff04275de8af8fcdca53591bcd5a34,8cb88e9cb559d8bee2b18e864f566c75e2ef8e23a9cc4dfa4904c01d996ce0e2,21fffb98ea7e780900cb75cf671674542ee9dd3fdfe75569f984945bea08fc71,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,5.42857142857143,5.4,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,arc-agi-1,ARC-AGI-1,abstraction,fc66952f5de68732185970ad1fc735883f4df60d1898f4e67f6cf82e1fb2ae3a,66ac94c0af5ae5438b8bf3438e0edd3df5c9e47c22ba4c3cab5aacd618a096ad,44ae69528a2232914d2fde6f474b0aebe672c3d23dbf8f660231e00d4aebaa13,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,87.5,87.5,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-08T22:23:14Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,swe-bench-verified,SWE-bench Verified,coding-agent,2852314d9246a3f3ce7d11d26eac27ed7158c68d99f8713d573a1582d3083daa,29713d07699290cd220bc92b081085a132d7679888a9c75fd226e93f18161ba7,0429901a09da1bb12ca4e339fa4e0d75ce501271a46e45422c8f87881a41a015,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,86.0,86.0,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,deepswe,DeepSWE,coding-agent,270fce12814894194fea42642329aba5171abf32533cd921bf9726121fa26c04,f4cd227ec382a01c5f7fd9b5b3525235783e443217ac6dabfaca7071d9084d66,54122d35ffdbd0fd3eb0527b7ca43a75f3d8c6d231a28990b2a760d842259ee4,release_12e8a62d68d933c29dcdf020,DeepSWE 1.1 vendor run,vendor-model-card-run,resolved_rate,percent,true,42.2,42.2,other-supplemental,supplemental-source,profile_f97c4e36649aacbbeb6c72f4,reported,unknown,,unknown,,false,2026-08-14,fresh,79,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-8-27b-card,Qwen3.8-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.8-27B
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,ioi,IOI,coding,fb1fd001743b5fb96397d2a9dd9b4f4dceadcb47f6b27ca0bd01e7b8f745cc6c,5009aed8d0d2f38516102be5f589889eae77c8f4d71c48be75c0193d8547cddb,ea360a73584f5f4a5b246ef6a7c00a9f4e4f74494a27db9d0261034e8d932adc,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,39.056,39.1,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,livecodebench,LiveCodeBench,coding,5202fd97b2169587ff1500d23bb14ef7bc92f78e4dec438d3b0f74d9678c910e,b8712f53448cfe12f112721318cc1c26fcfe7020ce9a8bc8c0f9864b55fbc6c3,a2cb431b496c9aa1a029eeb370c180e487b7c8d28a7a912a0e1139865d25f7ff,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,83.996,84.0,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,scicode,SciCode,coding,f7bb423e07bb7da9c4b2e0e34127dc9e6a56409b8574340f2059a9db721916eb,bfc9c905ba519e58232fd02dfd3bb83f7b60aad0c2c63f8862e6d01c9e6e9864,205a143825c138c40e6c26c54493d4c89b2783025790ca500e5c5aedd2858b5d,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,46.6435185185185,46.6,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,vibe-code-bench,Vibe Code Bench,coding-agent,8046efe303364e852245851e831d5a8107fe15ef3ec46340b830de5960ca1d56,74a318a31dd1cb81e67006157f874e064d897b953fdb005ec829df0be88583ad,a4cbec25a0618431f2bfa98de222cffef038f5a21fe946323775068084d5f73f,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,64.85,64.8,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,webdev-arena,Text Arena · Coding,coding,ed19049f9355143318a6f403247972b8e04bde770fc875601370d0b3c524c7cb,d73e68d214e7268a8370abee08c343f087ceea7d50ee52f6ff0b76a4bd7df418,6e42f3096498ca8ef0dc77728ec0ecb4a13898970445cc867abfd00cc8760b10,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1589.782268834193,1590,exact,canonical-source,profile_457b335e7d9b4a78950ca289,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,programbench,ProgramBench,coding-agent,9906b4e72b0681f54248c6df6a13d4f154c560a731c05fe0414eb5bf272df7b8,912dbb982b92a0145758ed6939877d130bf4994bcd33c88c0f1b9185232251cc,4cc623a277251397a3befa2bcf969d4beaab4a51795e84ea349ae44b2049f75c,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,890983ea53bc484f007550d98c54aa46a0e9010c4c5379dff7cff9953093fa5b,beb4f589c93126ede7001cf547b15825b3333232fa790d6096a654fd0f3e71fd,92a4752d2ba425ebfb7b606ec0e40266e013369a168f1ac90723b1505b3661c5,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,58.427,58.4,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,52cba7a5ab374e8635cda393e2251eea03cf2baa9749ac6cd2cf53235e01b59a,8144c97e658e0aaa8b7f8e635ef21f98c90eaf528bd0797ac57dc7290b9e031b,8c5e90a846f073fc49430001247dd9bc4b8be3e36d816862f9fd939f5f6685c9,release_b7d9ab744ffac0cb856e114d,Terminal-Bench 4.0 (Artificial Analysis),mini-swe-agent-66-tasks-500-steps,pass_at_1,percent,true,5.555556,5.6,other-supplemental,supplemental-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,ae5fede3d0ec3894df2d05e10e54646c5ce810ae113b6c2ae1b7f95457c502b1,4336c68f3498ac9e23591a49224c9f441487704744ee30ccf75eee49ec37e51c,fbacfe912d91aa95f2879238c50bc964b4a9d604cedf68fb0d6ddd99dbe61233,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,48.041237,48.0,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,apex-agents,APEX-Agents,tool-use,7221f9b7fe3c73a197cc12437c05926f60da2c629cbf96c90f6cfcf85f579812,1111d5bd36be5a4387f076174b673c330fcb912b809978d54bf25517286f50b9,8e74ff2f86dd73b1aac7df55aec635a29e92bfdac1841491c002b06a0589b170,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,47.5,47.5,exact,canonical-source,profile_bc57e8524f9d921805459b5e,reported,unknown,,unknown,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,arena-agent,LMArena Agent Arena,multi-turn-agent,66baf338bc9c7e033d8628964c290251e878c7016a8d879349168cdaa3923322,93357dff5bf2b0b058869d352f3b2b884e88e2f2ebe65a0ac648085d37854b0a,5b76e190acde094a3c38a328485bd4d2507b444537856511b98c48136301344f,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.04459740534602481,'-0.0446,exact,source-specific,profile_8d97cd82dc667edb7aa71a27,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,941544562df39f15f0557103341b1f148c35113f59c47863b5b20df5354eec6a,a55635d6b0c827b4dbfdf4d1862ad1aeaee5c84ddcf6ab3373274618be729def,7ce1a8177eeb6e001dd24b30122741e3332afd755f792975fe1307ceacfbb34b,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1426.73,1427,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
alibaba/qwen3-8-27b,Qwen 3.8 27B,Alibaba,open,55,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,d4753383f36a785d89352040cd7cbc406262440211cf6801491cc0a70dba7f3b,0f33c9d613fbc7c7fa410c692198cae57d7bd18d4975ddceeb98eedffc455898,381b4d985d09f53a654aa74d214a893d8c7961b171af26652e0ca59b4459f716,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,0.0,0.0,exact,canonical-source,profile_6037c094ef27884dccac420c,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,epoch-eci,Epoch Capabilities Index,capability-index,2e6258cdbf94960b17f474ad31c9fb56023af78b01bcdcdab177865882a8533e,9e646687393474ebc82b828673eec5119307231138361cfb84a08d6abdfee373,1a36a3e1a3920057423840e4e0f9acfc2f96262d9408f161902586823dc57d98,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,151.06,151.06,exact,canonical-source,profile_b310499954e2e03883d1bab8,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,arena-text,Arena Text,preference,803c667c2a6bbc253cd8bb48b4150c13be27e0824d4590e63b5a426c1a2230b6,06b3e33bc702f9ab7e4223803f589608786aeaa871a86642ce11d14938672a52,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1460.9562197110456,1461,exact,canonical-source,profile_efc5f41143198675a7b4f23e,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,livebench,LiveBench,general,79e1112c7deb5ec44228dacc38f068fa8e466cffe7ecc7e4fc6f424dca875251,e84c89ce6cd8ed3ea2ad537f45ba14b7f031ddc3639fb3cc640b7d5efc27ffb4,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,70.53878571428572,70.54,exact,canonical-source,profile_42ae63213d558eb4662e56b8,adaptive,adaptive,true,adaptive,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,e4bcdb2641b42e6fbc95ff19da97b21a0b778ef140dfd99b395e50ea1c2efe66,ec8de890b40885f2460eb1bd33be7a7283c41031722853f16d9a9c4f9a491dc5,fba5f6b10977c74d12a0a51f51615d7e5547d98b2fd12ae73799ca2e93a9ceeb,release_5f3023448c2ae4632fe1a43b,reported-runs,public-multimodal-mixed-run-configs,accuracy,percent,true,29.9,29.9,other-supplemental,supplemental-source,profile_68921aa42095f28a6654781d,reported,unknown,,unknown,,false,2026-04-20,stale,195,2,B,2,B,1,A,evals-report-official,0,evals-report-official,evals-report-hle,evals.report · Humanity's Last Exam,https://evals.report/benchmarks/humanitys-last-exam?tab=scores
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,simpleqa-verified,SimpleQA Verified,factuality,7cd9a373305ea666cc786f7743848a194492487316e10ac3b4a8dd7aba67e91b,b68e02849ecf09e56443eb3ec156b0c5cdf69a24b0426a037ba2ab9f465696d2,09143fdf56d84fdffc3344df3475d5a69d20be0048770624622b4be3711fda75,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,34.9,34.9,exact,canonical-source,profile_97380cea8f577d9e27da9479,reported,unknown,,unknown,,false,2026-08-10T21:35:30.000Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,gpqa-diamond,GPQA Diamond,science,6eb01a2b3b4153b382e044ae8d338e9bbe8953b3b736ec1402c439b21aac5a56,9f42954e1082e7b108d9e5e1f1bd13274cdce9414f480947c199ddf336ed7eb2,5092b11a27bd8a1af615eb90534dec811fd0d5ef01774616ac09dc220634a8af,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,89.142,89.14,exact,canonical-source,profile_34a077e25a13de6727203c0f,reported,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,mmlu-pro,MMLU-Pro,knowledge,b90e3ebaef417db40fd44f219c2f439e9a8f5d3e3511c2e0f20c894018add995,ade16b5f83e8e96ac3f2daec22784e00a6cebd3fe0b6acc0c343273db91aa273,a91873dc75c6740e34aea668d484eb2471a53fe70cda740ea02501315a6b9962,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,87.572,87.6,exact,canonical-source,profile_4b0cec355eeaeb214599a32e,reported,unknown,,unknown,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,a9ff6be6f8ebe3935cece7b53082635b9affa1690e1bb3fc8e0f70e0fa5ba287,42bca31c9830028edc40d13ef2d2b0a0b2be219cfbb1a5b411274cecaee6fddd,eee7ad52081a22a866018fdb36a96a8d45cf0b2c81142142245a800fbef0f45f,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,96.11111111111111,96.1,exact,canonical-source,profile_13dc2af2a60f2cc07172f0f7,reported,unknown,,unknown,,false,2026-05-02T09:08:49.000Z,fresh,182,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,frontiermath,FrontierMath T1–3,mathematics,a2be517f551835c2975f7376a4e6e9a7992011168bb71fc433cbea9a33cb1dc6,d019c653de1d0c6eab0205d88eadb08c01e9f89665b550ee6dfd1dfa9d17eff6,fd03000a890c14001f3d5c700ecc96f6eb91519f1ac9df62125290f68ae9c967,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,57.19298245614035,57.2,exact,canonical-source,profile_d1fab5b0d9cdb6adb08d2014,reported,unknown,,unknown,,false,2026-06-10T19:47:36.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,frontiermath-tier-4,FrontierMath T4,mathematics,c19bf8cee8b68ca4ae85854a900e1fcdec4f80244e2b0ec86006f482949ffed5,c330f739a1fa78daad33c3b1667c0e06d4ced91e2f2461d469fbaa0d8ae7a78e,79753c91c17a8c32905f7dda5872e1340706d508f05b9928e1eb76a02c7431dd,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,25.64102564102564,25.6,exact,canonical-source,profile_20fdd50658cc979ced3365dc,reported,unknown,,unknown,,false,2026-06-10T19:47:36.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,mmmu-pro,MMMU-Pro,multimodal,8da7a219fdf10359d80df463e888bb6ad42ee8744afe5ba156a100a7484b66da,cab689ea84ad923ebd18cb1a73138c40d1562ca98728e139fe587c3f524da377,67120e7fab38a96fcd406c42dac9466fa1d2b08f0b04664f3b763bf476380dec,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,86.301,86.3,exact,canonical-source,profile_f9ebb9b70f07c7fd44aa7c3f,reported,unknown,,unknown,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,critpt,CritPt,science,a4c0f43e613a6ba215b0e7918a9de8ce7698fb9469bf826e16ddcfbb96222a56,0305db3acad9346562b82cfa7afc5b7cbd83279fdf5e7a4cafc32428e3cab639,1cfcaea211044e0eecdcb06ccec090730c95c6d8d8d3b891766987ecdb5a5c22,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,8.0,8.0,exact,canonical-source,profile_272add6740d961e9d97d445b,reported,unknown,,unknown,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,chess-puzzles,Chess Puzzles,games,5300f83b048cca3903efc3b3e1e4600844291eeadf32757593ca1ec49297c5b5,4f624dcd885e0807fa4c819b5006da66fe93e1d768b676e3866668cb2247c8ac,6d9a0a62e0c88726d5c842123588d3d8d828316157ab166b6e89a2e305e5cf46,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,20.0,20.0,exact,canonical-source,profile_5933a601e7c8c1587b769e58,reported,unknown,,unknown,,false,2026-10-13T17:16:43.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,swe-bench-verified,SWE-bench Verified,coding-agent,88c432b7518a7b204ef8e8757846e901f39ee473f968b2b856dda225cb723e1c,175c3aeb6498201a97e0c74e949527fbc3811560a4e1a2923112fcad1e6c952a,88f8440a9094f80a17fd6eb95f51ec4399a09a41c92056772c872a6dcc15adcc,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,76.2,76.2,exact,canonical-source,profile_55a0f6cd9901fb49f5645d5d,reported,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,swe-rebench-v2,SWE-rebench v2,coding-agent,b21ad5872a6e89f68f9d0b98ad4b8b4d9720f2264d892a74b71629c9ab867ec3,be694bdefd40775a51817ae697c4f62023561dcb56ef9ada9173e1300a3291b2,4d1be04b07e6b95bf0f76baf433081fb6639c9fb4e627252e3d479715ccbfbaa,release_72d764965dd7baf34f983249,rolling-v2,2026-03-01/2026-05-15,resolved_rate,percent,true,46.54545454545455,46.55,other-supplemental,supplemental-prior-window,profile_85d245e50cf88f18729906fb,reported,unknown,,unknown,,false,2026-05-15T00:00:00Z,fresh,170,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,livecodebench,LiveCodeBench,coding,9a7a82ec9015d77a9ddc5f3dd005ecbb9ada3c69afb6b63026cf4383a1d7ca6c,a34028f9b621a061f50d650e30b34a2a893436de3f1970af28305c47f56d8f1c,400290ab721be6142e63bccc9ecc1fc530128e950d27e10f7141441d8e14d025,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,86.771,86.8,exact,canonical-source,profile_0921b0a660bb2fa8eba5dd75,reported,unknown,,unknown,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,scicode,SciCode,coding,5c89ffa04a95b2825a4df9758c78bdb5ac69f2a8b82267c049650d5dbf23d842,ce1ef162587cf6f237585ec94974062714093a54057e3916b560a16492fcff1b,c96e38cd745c305f2097b7074864adf7246c8a809c41b1eedb41bdabcef11817,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,53.4722222222222,53.5,exact,canonical-source,profile_96286f612a6bc490a7c6da8c,reported,unknown,,unknown,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,ale-bench,ALE-Bench,coding,8cb0ba7725fa149c884dabd7728e97ec29fca0689e138f10c4d2fee4722dfe67,f37117ca2f0b6dc5b72d57da3b2dcf9c43c67f072185dea5b7b7cf56a95d8af7,3703c854922fd87933ffac6a03927594645fb65bf266479163effe7a72cbc1cf,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1092.67,1092.67,exact,canonical-source,profile_b51293b6b43d208ca931a702,reported,unknown,,unknown,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,vibe-code-bench,Vibe Code Bench,coding-agent,e7ad1babb6553c01885c7a844189729153986852b18c57d54b5a5dee9d5d6645,7aea0b474b876075e91cfe70e19acbc3e6a397cbf592240bc8ddfd9b5c05a2a3,0577014ef287237c351705ec687359bc66a5fff7e5a98f55a202dfce4652778b,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,37.891,37.9,exact,canonical-source,profile_9eeb07a0de769530f935f98c,reported,unknown,,unknown,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,webdev-arena,Text Arena · Coding,coding,34683c305fcfd0e020f8d530154bcedcafbf2f323335dfdccf72acca1d36f4d8,f1cd5e27dc2fb17e6d635760520d7414a798fe5b9f0263a35e9734a351562032,b6d68f0e330c4c42ff79ec43a679019070fc63ec56d11461a675d60183ea159a,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1509.2234110707893,1509,exact,canonical-source,profile_768253bf922f44c1efe3bf53,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,programbench,ProgramBench,coding-agent,04c46149ee98c54afd91cb4ea0cb8b7fcb26eca64caa20719a68aa5b76fbd1b5,f221e1218b84ae441a54724172973ac40c587f052108183b9a341e19c462ba13,3f08659201588c1d8eae9bbe331be789d3372e8a2660ab9476713731388c0192,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_cf9e248427b06c9503d322c2,reported,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,641da7027ee22f84829c09b95c4971f4017e1dad331f649ec6427dfe891736a1,fcc4f8557538032a74a2e1b3fdab2ae50d289a0f2c4d79e49b4250c763aa69df,1eabcaa1cbc163be53205bae4888d46be72fc4139789f749e9ccfe9687560429,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,53.558,53.6,exact,canonical-source,profile_807bb82ee74af1856495dc68,reported,unknown,,unknown,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,skillsbench,SkillsBench v1.1,tool-use,e5407904f4372dbe8c907a5b04afcd16ae258be8fe44092098eec21c49a3c844,398a6046fe0df64a1b6805979de434544381eb78c043bc13dfe324816a282ce6,c2ed2025e534826e5e627446f4493a128269710491b35eb2122abbaf19d56304,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,54.0,54.0,exact,canonical-source,profile_c8716685b104fcbca109c687,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,vending-bench-2,Vending-Bench 2,multi-turn-agent,6d3e24c267eff6f5cc605113a1ad8412c3406b828f8ba8fb7d3048893219ac33,f097194891f7968099a9b36ce782725720db3f55a6cb7a52650f2067b65dd89c,7a300a06689c1bfa973d94d09dc598138039797d33b8ab97c0f89de019eee6bf,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,6204.570000000001,6204.57,exact,canonical-source,profile_8c49ed901c348b23f62acd44,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,d099f0dc777ae3c19ae2c847a145b399a8ebccf01e6b6af35b143669151ab086,e3015ec484dc140f9f776ae6aeed69d17cdd95587657f9247fdc17d72fcf6f58,75837e8a30ca4cc1a47fc90487485f460c44117608afb8622d6ddf8caa2b696a,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1046.53,1047,exact,canonical-source,profile_42ae63213d558eb4662e56b8,adaptive,adaptive,true,adaptive,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,osworld-2,OSWorld 2.0,computer-use-agent,0789f4f5a281cb92dce84cdb14045ab781dcb6b7841caca0231828ee82464a2f,90d37ac22d9518349fa113c8b00c896c0af070cd2e0b37665b9bf8ab1fc7631f,0d8f505dbc71ac537e0aa87c689661ebe93d7810bf90905213a1f0cdaccf5283,release_15c5525fe39f33bc181dfb69,OSWorld 2.0 current leaderboard,binary-accuracy-500-step-budget,binary_accuracy,percent,true,4.6,4.6,exact,canonical-source,profile_7a58f8d92e38556c6d31424b,reported,unknown,,unknown,,false,2026-09-02T09:27:06Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-osworld-2,Epoch AI mirror · OSWorld 2.0,https://epoch.ai/benchmarks/osworld-2
moonshot/kimi-k2-6,Kimi K2.6,Moonshot AI,open,56,weirdml,WeirdML,coding,9a7e46250ada3964ac8065fd0e4359604a65af626f10252c7f5c34ee74bc0baa,dd4c5ea3f1f229b96a86a3158de4a8955ab8e8f9cd1dfc4914835786d3e4c3b2,7c243afe23aa63e3900363e414753a05d10990b90cc42da92099149355363aea,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,55.86,55.9,exact,canonical-source,profile_3011d20cf3046664a1d12e36,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
google/gemini-4-carbon,Gemini 4 Carbon,Google,closed,57,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,95541d797511b058a4c245b11f814593ccb2f138ababfbcad40c22a92db46b37,ec6156b3b739e5da81da96fd4f0d554d4b3d9e9ef302efa8d1180910a7bc7283,c4d84d2dadbc521200297ea3702054b3edbfcafc1ba3a24f49b1f0a088f2c719,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,59.425739,59.4,exact,canonical-source,profile_ef2e61bb3d248820901f479e,high,high,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-4-carbon,Gemini 4 Carbon,Google,closed,57,arena-text,Arena Text,preference,4b1c6c03f4adb14c065a60d6ec174f4e01d7d642cf7ee6f7a786ffe5d14b1ab0,f003e202d7653e285697b4265cbe660a705b73d4926a3b02004cea6ceb81b4c5,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1527.5065447464403,1528,exact,canonical-source,profile_ef2e61bb3d248820901f479e,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
google/gemini-4-carbon,Gemini 4 Carbon,Google,closed,57,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,57d05951fe50028a2798246d0a8180b405419009e2e2b91e5bb8f086a3844b93,5d9a72461a39864d9a07cb6359908eb60c795ebf54fd7c069b08f88f27b5fcab,c5e5d47080878b2de6edb10dbd025dd5e5c4c534dc4eec492f17bd3cceefe25e,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,60.379981,60.4,exact,canonical-source,profile_ef2e61bb3d248820901f479e,high,high,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-4-carbon,Gemini 4 Carbon,Google,closed,57,critpt,CritPt,science,7f298c7bfb875e9bbfb473d6fad098df57aa5945f32dc7f0ce4a888f67ab559b,b0d64fdc0c57326c81464cefab90896e42e00fc7836d73ec7264537482a47949,34338c9b2fadeadb5295df7ba4be76e2c32f720f04e4fcd0f463dd86137201c2,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,36.0,36.0,exact,canonical-source,profile_ef2e61bb3d248820901f479e,high,high,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-4-carbon,Gemini 4 Carbon,Google,closed,57,scicode,SciCode,coding,142104358a3d37bf12cb0822baa88d7c1fe9f66ec39172c2b6e4e9114b6362ed,3a716e98a9576165ec5939f1734453c6785e3a19d7d8879e5fbc17a574450456,455dd738b5e3c71e92887fdffc46310d86b2b175c73b14ec36f5a5ac941d636d,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,64.2361111111111,64.2,exact,canonical-source,profile_ef2e61bb3d248820901f479e,high,high,true,fixed_effort,,false,2026-10-28T21:25:02Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
google/gemini-4-carbon,Gemini 4 Carbon,Google,closed,57,vibe-code-bench,Vibe Code Bench,coding-agent,272ff4ab2477cc7cd2697adc6ba167e03855d318086bd38541f92ef2c36414b3,5f3304ae158f46f8ade7c6995684a8df7ab353eea5b91a617eebce55551faed0,04e5aaf4f23019399011589f9c7a66735773fe5e181b0e9df63f715bbdaaa447,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,89.06,89.1,exact,canonical-source,profile_ef2e61bb3d248820901f479e,high,high,true,fixed_effort,,false,2026-10-28T17:21:40Z,fresh,3,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
google/gemini-4-carbon,Gemini 4 Carbon,Google,closed,57,webdev-arena,Text Arena · Coding,coding,163c196997bd7ce97f9a7c2d6025282f4112bfe4dafad8c0c43c044bb582db16,5c3ae235f579c38ad087ab27f6951386c2762e266d2fdaf08ad594180aeaf15c,39e78e0869e437de68dc1b47580a9be86c44e4a6e51d73a262c872e2e1fce40d,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1768.1522310766236,1768,exact,canonical-source,profile_ef2e61bb3d248820901f479e,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
google/gemini-4-carbon,Gemini 4 Carbon,Google,closed,57,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,fb5f2c55f8e81aa18b5823c39127d5ea04008b3bf8fbd1d529eee1beb7aef469,3a1054edbae3b889fae3871c001fc12b8bd4b23d556510c4915c0b0d06e31a11,6a086a197e82461d0c102a47469a3d683aa35b33bfd3b9a55cd8979eb5397fb4,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,63.131,63.1,exact,canonical-source,profile_ef2e61bb3d248820901f479e,high,high,true,fixed_effort,,false,2026-10-28T17:21:40Z,fresh,3,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
google/gemini-4-carbon,Gemini 4 Carbon,Google,closed,57,arena-agent,LMArena Agent Arena,multi-turn-agent,0e34caf709604edafe701c970aafd9e2ca36f30352d56f3a6a7ee527cd367cb9,2f5d752bcb748afa36bee16bb9a6fe192f6b99dc289e60e5b47668b79ca342e5,14da946bb1ef3a2a6d55f565b4355605f01cfef3a52d4f2f4592720af7e54015,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.16038738558073073,0.1604,exact,source-specific,profile_ef2e61bb3d248820901f479e,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
google/gemini-4-carbon,Gemini 4 Carbon,Google,closed,57,vending-bench-2,Vending-Bench 2,multi-turn-agent,91da289e27fbd14fba0c714f19a2d156ea94cd687928a8582627cff0fec10ca9,7af1c291543cea1ba9213fe641b804733fbfa63b0002d20672fc013c941395a7,d407ee77c4124e37711dd76fa884be59b9032e743580984a40b671e739b0b306,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,16484.20666666667,16484.21,exact,canonical-source,profile_679b802baec2b647dedb9379,reported,unknown,,unknown,,false,2026-10-28T21:25:04Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
google/gemini-4-carbon,Gemini 4 Carbon,Google,closed,57,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,be4ec243000874f8ab9eb312e6d4f49b269108308039c6d26bfbeb35b8339325,42eb629c537fddffae93927b72831e39380fc8cfb3c9560c644fa3adffedca13,1a5f92827d69d1b1f5cbbd17a7908fca6238ba6a200f877c4a06ca8026a1d323,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1732.68,1733,exact,canonical-source,profile_ef2e61bb3d248820901f479e,high,high,true,fixed_effort,,false,2026-10-30T01:23:12Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,epoch-eci,Epoch Capabilities Index,capability-index,4970afc68c0db5721543b8fee93cd3e08f9edcdd741a8e2dd46d5b6d08bd4a78,721c518c90da2ffad07dcf3118582e224ab89e2c20c0e31370e78d610a3953cf,b8b550d08d364703b48b080e40c5e90d183330c9191a02fba5a517ddf77c8534,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,148.51,148.51,exact,canonical-source,profile_57fc6d9a28dcfc60bef75504,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,7f52ec3982ee0a999a9b8f433b643cf525623a7b3abd210a63fce94258e43e16,064d3a423182585065566307fdf86b48f70f1e859cd2ab02b813085610f4fda4,63bf02113ce453ed20f8747588b2013ddbf4707d2cb68bb4e7a10a6436327d0f,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,12.47817,12.5,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,arena-text,Arena Text,preference,dcbeab653adb2d0a9ce0b5c7d8de453f1ec00f3add5607e2c86c0b016b6aa5d6,0b6602182eb8858b634d7c475a32ff07426bbbfe4158648c2ef6c7e5876fddf8,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1442.201111233133,1442,exact,canonical-source,profile_b73fb4763ea0cafd73adca99,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,livebench,LiveBench,general,5a68deb6e085d52ef45610156f2b52f9caca19bf2112b201a42dfbf6642ddf57,373a470e46dac02c078d8990dc62555990c41d8bcb2c2809e0b9e922184f888e,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,71.92241666666668,71.92,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,25bbc941cc962f4b506ba064e1cf5148c1ab07a6707d22e75e7481fa545ee313,28ace341855c8c473f37c6cf975f82b24ae926710c03500428512cf5d1c2574a,e1f956386a97e6ba42a3c396757902d19f19cdabb10fcc91e345af5c771a2890,release_fa9e908ac04b067eb00182d4,HLE official leaderboard,published-leaderboard,accuracy,percent,true,46.0,46,other-supplemental,supplemental-source,profile_9917693906cdf9a9f6c99327,reported,unknown,,unknown,,false,2026-08-29T23:52:43Z,fresh,63,1,A,1,A,5,C,self-reported,6,self-reported,hle,Humanity's Last Exam,https://huggingface.co/api/datasets/cais/hle/leaderboard?limit=200
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,53c786ce63f4dd0da8e801dc71006cb2ac13cb59eac11086802f66c87b2aebc5,39fe55615877fe5d90f9380450c5a3be173813f0bcf19708bdcad4f901723784,7f0862a25debfb669368622a3fddefcebe68ea6f18f784f8cd24602814d640c7,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,31.881372,31.9,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,simpleqa-verified,SimpleQA Verified,factuality,45f5a3fb5e23060d78b5f26f6b81804a2b7623ce3744e57cbebc2f874be5977f,9b609315dae0e488c10d46c22e58bcdc762932e9b134c127d3c6c5cfe630c13e,20eb758f84971175d825a75c364c42137740841325eaf7993f5a8ad3ae1df78e,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,40.300000000000004,40.3,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T19:36:45.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,gpqa-diamond,GPQA Diamond,science,6bb8ca8d57c2a563c1ebab41e832badcc4a107cd7f707892715face7abe7549b,0f1dc49728933183dff43a65eaba790b1886dce37e6c837c3a731e0194b1caca,eab2240a7ea2afea2366cddd07324c0c3addb6191e64983fa6e9ea882744a579,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,87.122,87.12,exact,canonical-source,profile_7279a3f7a28bd9930d9b03cc,0.99,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,mmlu-pro,MMLU-Pro,knowledge,580c61f205eab92cc0bd1e0979e6960c1d5239018d8cd91250411adc1746d5cd,dc7db2cc9ce053eba8b01ab8f2caae0ea4ab154bc1ca8c48bc4b447c9646a5fe,7f37d3be152bbb7073099c5a5d7650b6b70a0e912c3c111ae743a63ad5fa0955,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.295,86.3,exact,canonical-source,profile_c8d9902582e198f49805a576,0.99,unknown,,unknown,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,11810ddf46ca924536300ea312ffde64c4cdbba0e4bbe98539aeb020db8a3ee8,12b91534e2bde8e1dfe2cf67601c1018c074ccabc4ecfba514ac966908b6e07c,d7ccddb87a309ce5d6c745f39bb12910ef2327bf966e1c993a0a4c524bce446f,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,88.88888888888889,88.9,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-08-05T18:55:39.000Z,fresh,87,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,frontiermath,FrontierMath T1–3,mathematics,813407a7c4d8e82989ae0ae7805d5ffeb9ae752f9781dce82953aba9a138c151,e4e24b73639de33451ed6ce5179ebf4a73b6498fb08e3ca6b62ae7758c26fec3,71b356cdd9bc989e02c5e1c7ada49f7370ae5ef3e39de146b938b0bd88eb988f,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,33.33333333333333,33.3,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-08-06T20:14:32.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,arc-agi-2,ARC-AGI-2,abstraction,fa687947c022d0f1e6d436c7c2ac02664df3fefa634739de73390bdb9d592a45,f867d506fc1010a5fd4e433e6a3ff5dac2af9a165114f2b971e8c92c917beeb0,84ab3445a8f1e639d5912777a39d57baa63d5da37edee2657435ca577434367b,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,36.52777777777778,37,exact,canonical-source,profile_9f4576349b0b9a1dcba8e325,reported,unknown,,unknown,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,frontiermath-tier-4,FrontierMath T4,mathematics,6bf6196ff5b35f17817d5e5fe674717a42bb5ade51797f2d565e6cd2e06981da,0cfdec21e3fb756793e7f86c2f7966f80fe267b47318badaa7126de325d8e4b9,5d06392ee872e180a995f19d965db004f09484e9081d09704fbcf92473a6e0a0,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,4.878048780487805,4.9,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-08-06T20:14:32.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,mmmu-pro,MMMU-Pro,multimodal,8878841f72f67ebbfb49be1f59d9e8344b3bf26122cb62122c814743c2f1b1d3,fd162a4198db280d04c4ff3231d51faad58c18d2c628983a5614f26e78d45b22,5d10cabb7b749a985b6e91c0c5a8ccca1a2efc80b4461f90ec7b42da9d0ca1ee,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,78.497,78.5,exact,canonical-source,profile_fa698e6a574cf4d42e206b07,0.99,unknown,,unknown,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,critpt,CritPt,science,b0497db5e1a0b577740335178835abb025978faea8d5566bff669e8cc15b3fda,dfbcd9697cedab6d1c5a87110e99855fc5ca34bf2f28fe272cba3b646692a028,ecbbe008614a3c58bf3e0adcc54d3a3d232f589b2a949eafc83f1644549d9f14,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,5.42857142857143,5.4,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,simplebench,SimpleBench,reasoning,47cd03116df620a196306cf8465da660397a86c06d712b12fba071589546b694,daeeed7a5aa5aa90073ff9d6ace6aa7164dd65eced31aae9aa0a3b2dd7bb3b14,f6b61380000b7f489f75162e5da21a7d50a80ee616f5a65ccd9f104704b53447,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,50.0,50.0,exact,canonical-source,profile_323430f8d70d985199accebe,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,chess-puzzles,Chess Puzzles,games,b5f0d154bc8448d87c6c94146385a4c7071acff056007d85e6e904b70f1f7599,2efd8dac71f20bf5948956165734fbe3504f8ed5ad47b541753bbb0b8bcd00cc,d18b2698e7fc230e2af92daabb030998e3f0f1722bf5cff30628c76c41c44052,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,21.0,21.0,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-10-13T12:49:23.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,arc-agi-1,ARC-AGI-1,abstraction,7697b7a2b2b14db76174547bbf9c821110e4ab35fff84baaa1ee939f8b5c9551,90d7d869c6e1c6d4e29e70c66803fd6041b56287a7f7d6777498721d9113a584,293efacf877896e53be0b51a4aef786c41496191bf99b1897934d470ed42dd8a,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,79.5,79.5,exact,canonical-source,profile_84441dcf4ff9d3a7ebb4f5f3,reported,unknown,,unknown,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,swe-bench-verified,SWE-bench Verified,coding-agent,0d2105fc8657e6e7a3e7934facbc566e62a1f299bd9724038f19b12b728f0733,4292f39bcdb31e7a53feffa5304dab99f532a1e995079a5830ee695a17d9de4e,178ad2c09e2fef8871c623dd837f4ee2979644bd9a7f685a6c5a2a34dda2a40c,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,77.6,77.6,exact,canonical-source,profile_20afbe6979e993a50873f13e,0.99,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,d2362316f9a0717a97c36b8b83574fab69b93f8afa2283ff8fe3b2799083b5ee,6f5195cc5542206c76167050787f838b364e06fdf0162ef97b84717dce012177,5aaeee6a3785d2c915e480b25e8a0c6e59f71d00a3373aa927b0e3b63861578e,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,89.88,89.9,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-10-07T01:44:58Z,fresh,25,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,frontiercode,FrontierCode 1.1 · Main,coding-agent,31b8c188c5b2e3c1f0486616f4f5be6a0d8270c5cc9cca4e3ddad256a30024a6,a0263a3ed5e167984b1ad5baa78aa3e8abe61902934e44bc3d403333981b2ddb,4f671b8d19afc203b8b3147d4f5f2816d40402fbcd2c142155ec201ac578736f,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,14.000000000000002,14.0,exact,canonical-source,profile_b8c36c61e899913b2bb3beed,reported,unknown,,unknown,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,ioi,IOI,coding,6f2d2fd8bc922eb43ef1f1c5f7e341383c52289674cc475839d8c9f493106acb,92165c11c71cfc7a4be9e5eaa1ff7fa4d1bf90a47bfe9ef480ab0e288e459a09,a1e509f0f8519979ad82270626dfca0ae9e6c4c815f28e97f036df46eb00c637,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,14.944,14.9,exact,canonical-source,profile_aa19ff77759412701a03b4d0,0.99,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,livecodebench,LiveCodeBench,coding,4500f560f441a70a909a984f5a1b5a658e3a3d5de1708b0b0fa25cda28f40100,393f218809d3b79e7e88cee07993e04d02408293290c5bea7ea851625ae76a08,2fc824df4fb3315c29a1859123eadc86dac59417f8b7b09bb4fd7e9f2d3634bc,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,85.517,85.5,exact,canonical-source,profile_d9ea177d067409c4e24792fd,0.99,unknown,,unknown,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,scicode,SciCode,coding,795ba6fa840cc52c51ffef54c95e1759630ad0e56f451716c8af937d31fc9d9f,2415f6ecfb637e7d451eb58c52ee18e30f06d9d0839e81675054c10d1055751c,e70a081ece68126afcf14695488ee1d26d6ce32b85c428cf6263689eb284c447,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,46.9907407407407,47.0,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,ale-bench,ALE-Bench,coding,c7b80936d21e5e777731ffb7a6b34ab555fea4b2ff2060205331d0b39d0720e9,72d0f368e6476abe55f21c62000860db5b1c2d6f27fb52a47e8697949b9ef2d8,15a3f86976f978661efbb94596401133b13178dd3ffac3599cb852845d2f49c8,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,946.0,946.00,exact,canonical-source,profile_5b46a048212c2bd56946fe13,max,max,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,vibe-code-bench,Vibe Code Bench,coding-agent,b7649608d2b6ff7102066d6dc7c6073eda33efd7987a3b9b891ab7f7ed9aebec,3adbc321cc829be29be99496fb6bfeb4ddcfd90705ed7a7f571a6b4b4bccc003,976d585baec8bdadde672f17193cc6586f99100d0e4cbf61345ac0a9c1132a8e,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,19.214,19.2,exact,canonical-source,profile_777456089accb2a956038d7c,0.99,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,webdev-arena,Text Arena · Coding,coding,6f4f36b1b98c760c062e205d1c0b1a8d04eabe2784d5f4abed7f3e17b8bba505,597952b03529a149f8fed6cf88228499402ff9d6bc3b5ddbf4603f0d7d79b0dc,ea835c5aec7baa48af8ad688ce932d00eff614eb4a4fd9478b4b556bfcd4ccb0,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1415.192007852744,1415,exact,canonical-source,profile_75b2751fe3dec6d639534bf1,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,programbench,ProgramBench,coding-agent,197ccd8905f941d98d13c5d1cea099801d0d2180bf94155b484789ca391cfee1,36475da28262580a0233d90d43cebafb1fba39c5bc484b04b14e6e514f7c83f2,f5cfc600a9accec5d0ff8f3059124bd2410a94efb6d16aad80dfea4bffb92069,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_9fb27134891c2a3b66e9b8de,0.99,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,weirdml-v3,WeirdML v3,coding-agent,897c8e6468e6d583cd3ae772299dda9a4067dda8505a6876a593c4fd11907dbc,1d96b2dd541c5cccc370af077cfd9a9c1a4802cfecfc41c171ff6f58ff0ea610,13c017b8386b2cc196d73c6646ef6de9eed9441b45c56254cb57f018bd2b43c3,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,3.9029836363636363,3.9,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,9d346d00f9e78d8a468a3b3f224957cbf381e83b0e4ad99c27a8e143a6167b12,63da7f5dd10c6a8ef2c71b8a7ee4d3c85600ebc3f6bb6a0ba2d1a5589f05dc2b,cece4b3d8227515c03aa4c5d4bcaee43541264ad81315bc7fa0a5d945676a184,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,47.566,47.6,exact,canonical-source,profile_551396cfb7201bd63ba1e871,0.99,unknown,,unknown,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,4e9cfdacae80678b9ce1d8904b5c3701abfb838806c210d9116c708e9a272984,466c0140146a637ac9e56034821f1fe9d37e79d0991abcd6e1977ff22915a941,ac8ae1b27815c194fedd262d4630054bce0548b3bd7b11dc4bd5079d7fe081e0,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,0.505,0.5,exact,canonical-source,profile_f608f4723bcb538ddab99191,0.99,unknown,,unknown,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,sre-bench,SRE Bench,terminal-agent,98446077fa6a51c0768fc10badcfb16562e7faab6a5077250a18e3c1018c044a,70786ee90a103dfca3d7335f3ddff1c4b978c4495406f9968e0853ae12ae5c4a,91c5039acc22e714d75ba3a69cce7c81c4de69debad4c2fb6d51da9496d825d7,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_8d3b7dba49e2253ccabbab53,0.99,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,04f6027383f49c73024c3596e432d1216593bbbb81d1886df5a23bebfb211ea8,47f990a214a5b191c5c73aa41d0a15033cf53df42974d481842ed6be4c0ec496,b9dc6ce08f2944cacf598c9e72bd288a7a4376e850c70ac6cb563f0fdc531934,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,29.072165,29.1,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-10-01T02:30:07Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,mcp-atlas,MCP-Atlas · Scale,tool-use,b882166b1876e600cb2eb5a88b17a96760aa8f949229d94b1fd5a67598191126,a4e1dd61ddbd8f8a5d7f64699108fadafc9096fb5ad2d4b6ae26d23258a98d18,4dfedff3dee4ea68f6215bb658aa8d460685181b9524219f1be6c471a30d2bc2,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,76.0,76.0,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-07-15T18:19:13Z,fresh,108,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,apex-agents,APEX-Agents,tool-use,c7a4041776299f234e22e6962c6194921b57fed0b9b14cc7f6cf9d12781e3bef,c48f2a911700e701512bd5964b955c0dc24a591485f897276715f9386be2f1ba,271aff4c2e1b143be813fde8fd1c641c2f10f92b540154a47c713fef5707ea2a,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,33.8,33.8,exact,canonical-source,profile_2c3ca31d2106579e73b85c2c,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,arena-agent,LMArena Agent Arena,multi-turn-agent,35fa3f469523b52e17feb29cb0f5472051fbf25630323ea90599ed2941cfe49c,54ee42be05bb2a35f076b81a2ed096aa6b9cc39767a89ea3c5e58f472a8102e5,0f592ab5858e3b737b1549d2b862d547e292e081524561154d721017c725443e,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.15799537295041588,'-0.1580,exact,source-specific,profile_868910b5371c3f0911dd2a6d,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,1927787518250ad46bb54d258515d9280773330259123b85284bf44c0639d264,debc15d4ce52e7375f97ea991eec24138b0739727f33409d3edfb6cac1d470f0,28115deb8d5a4ceb421a7ac3621b3893a92e169718e0a321a68cb61a2206e9ff,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1080.22,1080,exact,canonical-source,profile_d7ea10543bcfd3f2dca6a744,xhigh,xhigh,true,fixed_effort,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,48920c5e32ad3ac8b385acf9fe909cd110d596fa99a69006de811b8f925af3f4,44c4233e11067047bc9a5d2057a2bf9e49db4a79b9bafdd5da2559d705774126,9b3a9ea0b94586db52813b863262b4d8c3a015bf4bc6452b53ac000c57e1562f,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,25.0,25.0,exact,canonical-source,profile_5b46a048212c2bd56946fe13,max,max,true,fixed_effort,,false,2026-07-24,fresh,100,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
thinking-machines/inkling,Inkling,Thinking Machines Lab,open,58,weirdml,WeirdML,coding,0378fe919d6a2ca882b55755d176cc48ecda9d47b9e82fc5a170077374b5edf4,f029aec3d1ffc183f6279a649e177c101bfe16668abb5dea859a02ab76649902,4640d6de029d479249606816cda2c2b788d7aee89ef4aed6716d30abc6d774f9,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,32.29,32.3,exact,canonical-source,profile_a026da84feee358d3fb40d36,high,high,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,epoch-eci,Epoch Capabilities Index,capability-index,6aadeb5c81ef0517fcbd0b55169981144d768577542c55d03e474552e3853057,8373ef2208842174d1f2b684816f0577712b8304830c548bdd06fd15fa460c0a,7f7ab96d5526b92002f36322aede653e346e0b9d51045f17b1fbfae1df555c48,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,149.13,149.13,exact,canonical-source,profile_2f74bd93cdb08957659a5334,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,arena-text,Arena Text,preference,da8b49ae999da132ea7181cec054cc839593ee172bf39e0faaf701c6486ab10b,1f53a396e6ae0dcce249d880ce6310365f8359affd8a40c6ddbfbcb808c84ca5,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1454.894064898029,1455,exact,canonical-source,profile_c5963039dee3d753db6b264c,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,livebench,LiveBench,general,1d5a6736128d08d1295fb397e9ebb37ee57cb0c83f9f4ab94d697914c6add77c,e6055b2b9693f472f3627928d741fa26ac948235ffbf4780f370302dbd283d58,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,71.57225000000001,71.57,exact,canonical-source,profile_192f474c0dccc604e621ccf3,reported,unknown,,unknown,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,6092406147b9796ef0065ae304dd37ef4fedaed10752880958e00c544e23de4c,b4dd9bd2014479d5ebb7c2aebb554c42f342c22bb4d5ff2ba7cc81440ae450b6,9d79f00315c23bc47d2293c9003ea42de6ca857854c04113ff97996ca36b3878,release_1f4bfa5486eb4deec4942449,HLE original,provider-report-no-tools,accuracy,percent,true,37.7,37.7,other-supplemental,supplemental-source,profile_79ec886ae22fa1df9c9fd7c2,reported,unknown,,unknown,,false,2026-04-24,stale,191,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,deepseek-v4-pro-card,DeepSeek · V4 Pro 0813 Model Card,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,1d23da49e8046a9df0f84e88cfb40a78fb3a38158577d656c523c61344d1b560,4e434f129c7bab7dd1ca8b5bd1679b936f5c34d5ec7514bf0c3117b0d1a2cea4,a174714c0c02664b3978f4ef91c041cb4ccea658046754a3e785e15abc4b8e91,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,37.534754,37.5,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,simpleqa-verified,SimpleQA Verified,factuality,7e1baf3bd6de10d7bb7953a2301e50eae6c74fb7602e3eb090aefeb04067a182,2f1b468b12ae6f27ab66888d88e25a81821f70618efbd4cdd32aadafdd3c025e,e6b53f7246e0f1d6038e6ea750f2d839e2dc7c2c8caaff3af680b5977c371cec,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,46.9939879759519,47.0,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-08-27T19:30:29.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,gpqa-diamond,GPQA Diamond,science,ce32f7ef3f87be6f6a41aae76001f65637752dda876d45f05fcbfbbc77321faf,ed6f73b9fa1d0dc004528677f833537c0ba71ab75fd7baff0293d654d5065418,f751a91ecd23c5b2b2c3c92dca8a7ba18c596d90a35bd4cad13511de7cb85717,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,89.394,89.39,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,mmlu-pro,MMLU-Pro,knowledge,73044192c7015368816827af04b1a20abb5e4848ad16c047ad9335d3565123a6,59059c456fbb5cd9409a98bf884b69f291c50e88fe30505f385cf84c3bc070ad,38d09e0ec2ce46b7cbfd13c64d55be7fed3197482a5e4956d986d206031288cd,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,87.249,87.2,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,e5928266f7852511312d473b0809883fa143bb415f2201e3d8a3265542756aae,04c1162edd8bed35b4e2ca363e830ff8cc6a1e54b7e79a0991d8428307d7a1bd,de87d7ac7927a21c72e918946bc37f3a2db2ead14e8fdf928a8d6809fe55339e,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,96.66666666666667,96.7,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-06-17T00:34:52.000Z,fresh,137,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,frontiermath,FrontierMath T1–3,mathematics,5d498c76d15ee9dfd9a68b1ba7bcfe05ab156764bdf321fb6ebfe746b9723ff1,7267468b261301d8d37e7a26718b5abf46d372f73493c66ac7891cfa0e813f9e,95bf319888c76b6e404d381ba4159ef1791db4120bdc0eaeef0f05a4a760f22e,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,45.26315789473684,45.3,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-06-17T02:30:14.000Z,fresh,137,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,frontiermath-tier-4,FrontierMath T4,mathematics,5ee40aae447f0fa895b37da6bcf28d7685bbfbe8960f8f129bf79dd270a02507,d6ad453ea1bcc4e8f6a8b755ff7e9c4b110f13a7a6ca2cc9757034eee77a3fb5,3b144d21098efd3c14d0a87584ce1cda445178f5d3330ba826845d01323ddc6c,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,2.4390243902439024,2.4,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-06-17T02:30:14.000Z,fresh,137,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,critpt,CritPt,science,736f616b722467f3e2b057bbcb4da4db669a68172f423a8981a06436ccabe6bb,02f37bee68b6ef7bace6be03e07d46d8ac594841ffa32ead246cfa6bab99bd63,54c0a3462b0354586e63990949d9c5451c781ef12ee6b0a0886548f23f1aaee3,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,12.857143,12.9,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,chess-puzzles,Chess Puzzles,games,3fd24f816c6e8d82f8ed91e92ec62474560ae624b28c7ba70284e33f7ec39bab,19393e2990b77f0aa8b8ce7ecfda7e7c400c39839040f007c3999869ccdd66ef,bbd6f6993516ba2f74604dff73de67057cec18449049bd2a9647e0a7681b6f0c,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,19.0,19.0,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-10-13T13:38:19.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,swe-bench-verified,SWE-bench Verified,coding-agent,b3c879f19b1e2f8edf6ba0ce601052bdcc55d5aa10a592c16512aed2786e98af,e0a1f5d3fc83df22efcfc6d56efb48f70cc5b118ac32e938f6708d223fad31b2,14260fb5f15d155695a133e4901b2b4c82e4066f6fe3ef5402847374213ea6f1,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,77.4,77.4,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,swe-rebench-v2,SWE-rebench v2,coding-agent,2a386f936fb7e907b8d80ef4a5670300edfd682c732d9ce28fc756773e7bf775,f70282efdf208348c11971aac7b8650fa97d785918364c8b52db641895467dc0,7e4a46183c41dd422edc5787a41225794e91e14e6c1831d24d9ddfe3a73ac2e3,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,40.18018018018018,40.18,exact,canonical-source,profile_c5963039dee3d753db6b264c,high,high,true,fixed_effort,,false,2026-07-01T00:00:00Z,fresh,123,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,deepswe,DeepSWE,coding-agent,48089981dc59f3219a5def5e0e003fb716641159e16673c18a65d380ed13a1dc,63044f08c99197f6025bf056ef5ab64da147a305b01361e7b0c1a5e2e3dab604,88739783113936c77f12a3e2bfdf8872eca3d657a584a15474bdca234b72bff8,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,62.83185840707964,62.8,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,frontiercode,FrontierCode 1.1 · Main,coding-agent,4b23cbf459c122afd40525b98e7f1950be50197fe4ef2d8ac08e4f5b044e2fbb,e70e185b893bb67012e94055968e529b2f4d8fe4fcbcdf5824fe57a18cac0a86,b672ef2a03c8dd178ee749c191c14cd37184eebdde01297553288af327cf08c1,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,17.64,17.6,exact,canonical-source,profile_aa69f796458792e2a6955243,off,off,false,none,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,livecodebench,LiveCodeBench,coding,45959b6255b787685f29e2cf4f0cf7e8ea027eaa402a12f7a756d1d3bc568072,29fb92e122606de2bbf54784464c16f3ebefe5853c47356d9bc5506b616d1e61,99a7e6922848de92555c58cd110fa72d016709410b5154ee4f0a5791f27c684e,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,87.484,87.5,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,scicode,SciCode,coding,966ea5cc387ba7722e42b7bbfc6d1f197d52c5d351d87d74f9287d584bc0364b,70ed966d0b47868295d781f38ab1385f0453472a57cc2e6e1dec129e40222ae8,0746c2a76e3dea8e9a4ba6d60699960731fdae218c6d5df46786a9a55f76fb04,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,50.0,50.0,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,ale-bench,ALE-Bench,coding,675a5693b0cb124c119dfdda61825d6851377f99127a2c69bee745aa978676b4,c3cdda862cbce0c13ac6aab613eacb4d41485b1ebbfbd304a7a0acce3b74b930,d65099c81612af988ffb7a0b93a5544a9a82eb2b87403f6854871af4995567ad,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1006.08,1006.08,exact,canonical-source,profile_c5963039dee3d753db6b264c,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,vibe-code-bench,Vibe Code Bench,coding-agent,63416ed09fa4a0b3d63623c9d1d8dfb362fd429b971810107ea35ee429b26ec9,6e95e1cfe0df0e5b2b1f971fb37f98356ef4c480f00cf5fb312bbdebc9c47d79,942c7679ebe94cd311c86e129f5045cb9879022d3505ad0ae74d6cd0fa6519f8,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,49.931,49.9,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,webdev-arena,Text Arena · Coding,coding,8a97a60584d9e099541d1c21dced8c851ee8e62a20456c77a93970228b8dab3b,59932cb9d4bef9a4bf8f1a63177b8db5e64a9fec4ded54261c058a6263eb1fc6,c555068f83103a9aaf1a353989816e340d859d20148171139425022f65c86c30,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1464.190999873057,1464,exact,canonical-source,profile_c5963039dee3d753db6b264c,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,programbench,ProgramBench,coding-agent,19a9f8c9c6afa70310be080bc6c8253cf3d2fb027a04ff520305256e2e139222,f266f6c8bef5cf1115f981b504d0c25513cf13581a1683df1188f795281acf54,fa377c070d5b851d90b06fc661f87949b0c8129e7750e2f1b2aafe78dfd9e7b8,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,0011f83722e8b526bb4285b89e0afb0134834b308c1610e18b622ba1de94a7f0,0a3402a0f6af39ceb131ef763d91a8e4fa15ca942b548a40b22aaeda2a5cfde1,fd9857580e7684a852b9c4cd51d7cf1024c6324c5ed457af01e971c180aad987,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,50.187,50.2,exact,canonical-source,profile_085955b3c249b60f983db38c,reported,unknown,,unknown,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,8b0222c8bc7b81d438625a919333990cd7b5a545dbe21f96bf618f06bcb75ac1,c9b25b7071fb69fb1a5e589154365b2c294006149e72477019670b6f7ebcad6f,fc6138e4c585334d03cfb04ee059f981bbc0467a0884a073890c9de7f15f32b6,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,11.111,11.1,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,f4b1ef89cf99ce54f6388e22d6c0b7a142d5d9a8b9255a8371c5710c9ba7e73c,a82b480f9072654d60a7ab5212f556a571f4db61cebef3d5eed163a646c6eaa3,7fd2ad946ee97973206f8f5078309493d26dbf48bc1a656bfe529d9b14c6ecec,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,30.103093,30.1,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,skillsbench,SkillsBench v1.1,tool-use,6d93ddb784e357cd7fa697985209bd1de059491dc5861e01d620343c9675031b,4602c43769bc08a8edefac7db918794494f31dc3f20c90be2c2a4a57ef633c97,c2ed2025e534826e5e627446f4493a128269710491b35eb2122abbaf19d56304,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,50.1,50.1,exact,canonical-source,profile_1710bb809c3113ebff53f7d3,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,vending-bench-2,Vending-Bench 2,multi-turn-agent,3bd11c7987f61d18e5ed6fd57891ccf438c443e60af1edaab578a748646f7c3d,2932defaf40c34eef689e18801fcb5ca16646cdfae5054b7744f9e9181678b85,be87d9ad247a6773fa090746ce641b5caf524039d2caa36fb2af7b17aa7b965b,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,3284.519999999999,3284.52,exact,canonical-source,profile_1408cbe9e98576abb4a0f36e,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,9031b229e2aa568fe6fd792d63c44854e5e6edc9ec1c151ef7ab34ec52f3f70d,cf14937be22d717009930a8acf9d76e28aec0c1d16939dd123b5ab7a90008d43,435b584b368291ac8d9559b181f365b1ddac7c2e48ed24d78572ecc97340d4d7,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1166.24,1166,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
deepseek/deepseek-v4-pro,DeepSeek V4 Pro (2026-04-22),DeepSeek,open,59,weirdml,WeirdML,coding,41e02799184ddbdcce60f6068ce4130a9aa9b4fbfb8c3fe6e626cdf408a971d3,d69ed27d90c386a632d3811ca87a57a82a4a683d2b69409b614657b008ed06e8,db710865a7051263576e4aa65d3c94cd9cd42c3f47757dfe56df0f4fd5301319,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,48.9,48.9,exact,canonical-source,profile_27a543e4d8b8c6f237f312c1,max,max,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,f6e807753d0c38cbce9c0c2e475396c4141ec089bc3ed4aae1cb4be42039b635,a77d97ae423dd92b8ce3b8fae56271439d651eb6e692e68b37e856e05d6eb68a,763dc2c67f30ae1fb4f6c1d6563d0879ee56ae53b6b28a9d8f46dbd865868fea,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,29.889783,29.9,exact,canonical-source,profile_ad5c4858e74f428a0183b9f8,reported,unknown,,unknown,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,arena-text,Arena Text,preference,4fa1bcfd69d69cb9a1e51325a9a2ceb5bd4204f478c076342a695208c799427d,3d0e798658e463ac5e83af7354c74e81c1b4da446babe17992b887e0100cf5dc,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1479.6978910104858,1480,exact,canonical-source,profile_95f69c97aa4a1cec75c671d7,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,a8db1e011bf77aa7528ad65873cf479626d03cca881be29586c42b397ca0bc49,1bd7f678e402064210c5db38ed83757f90fb994c45025a7e0f1cd9b087bb3e48,117ce075587de96ffde64dcc0c76c0833d042aa82584ba6a8126b118e3d66729,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,49.351251,49.4,exact,canonical-source,profile_ad5c4858e74f428a0183b9f8,reported,unknown,,unknown,,false,2026-09-30T18:53:23Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,critpt,CritPt,science,faa6312d3ef1b6699ef3c0fb8ca44dda523f2c3183b46d677ae0e86d137a3538,91f3d77004647978016516f20645f485f73ae85284256881ce74df625f230525,d9e8f6e4576dcbcd1b00c9e181cea4a25fc15c1618e0a9dcd4d2a309a854c86e,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,26.5714285714286,26.6,exact,canonical-source,profile_a6d859a7de2c648262173a51,reported,unknown,,unknown,,false,2026-10-08T22:23:13Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,ioi,IOI,coding,38d89956aa04b38b95ac91c4e3c71e06000ec2b915589a80ef7e64ce1fbcd130,d334ad93d4e46a5c70fd5a80ff69aeac6d021db6069e4a475488050502cdfb8f,02bdf6e671f2215b6850a4bd42453f8262c954b2b71ffd01dfcd81344c0fc538,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,39.333,39.3,exact,canonical-source,profile_9371c32d8313eb1b572415d4,reported,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,scicode,SciCode,coding,d05316dc1f55a472eab87bf527cee060c4ba853ff62c5f33e16b3dd3ba33dcc1,72765083df296b6b5fa7bc26e4b995b976e9ab9a462d8a33fe47bdf848939df2,1f06c1b7534c2fb7fccc428717b63eef3e18954e9a35b0f1a1ba9305a2e9434b,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,60.8796296296296,60.9,exact,canonical-source,profile_6130f112650f283957c33de0,reported,unknown,,unknown,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,ale-bench,ALE-Bench,coding,df1260bfe99cdbf3f116d236deb5ebae08add0a5baa611204b12bef6848c9ac6,0f2fbbdf3694d6dfa734632063b5b39d7dd6ed5b4e78f700b9ef6dc6ce620afd,a2b4cf473ec5223f4c0d0825b827ef697896485033d1a8d5ca4ade9bdb96770e,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1158.1,1158.10,exact,canonical-source,profile_e0935f7f898ad2b33c5be391,reported,unknown,,unknown,,false,2026-10-08T22:23:15Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,vibe-code-bench,Vibe Code Bench,coding-agent,1d4b85595546420fe29f1cd7feee1746b14417dc47167226df7a7a7297edc9d9,4b6160935bd56912bbfb9ca408dfc259ae5e363e709089f69933111ff451549b,7ecc126394a88bb0d462beb53f3a33dd51f9e121920c738b67cd8db108dbdbb8,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,85.223,85.2,exact,canonical-source,profile_7c73c8831eb69ca88b4cc5e8,reported,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,webdev-arena,Text Arena · Coding,coding,f6b2784520c40a28801846d80f67bcd087a712659582293aea8356ce37340137,035c38ce4db4e292125845ffed1760a256f872e5dfde732d31380f5f0ca81a51,a346f712942c25f31078e2be54957a4c00fe8b0835f63f7b22147453f63d5428,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1625.417548626546,1625,exact,canonical-source,profile_aeabb9ac695c1ca1e41af8b9,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,programbench,ProgramBench,coding-agent,912387123238d36eba8113367fcdb3b2d4f2d9278fdcef4157ca57b1547c79de,5921ca50ad5bffa49dc437426d41de92d8fa9666cdfdb46306dcdc47caf096ba,4963cfa30d7a2b41636db35e8b58dbe7462166ab617ee219f9baa95b3752f571,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.5,0.5,exact,canonical-source,profile_a4af681ed9f15751db717d43,reported,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,2bfd61316c3571c1465fd73c474ead0526521f32168e9810f12e95e0ebe8f746,a535fdefa8dbce5dbdbae78ec0d731c1c80f27fb31ae24c2d035936ee60657da,209c2dec022ebd3237bcf27b87aa0e21c2a2734c519a2693164ca1994e3f1e3a,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,67.79,67.8,exact,canonical-source,profile_83ef76f94df553775bdb50c2,reported,unknown,,unknown,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,3e31e08e5a78ec4e54eb38d2c34c039ac0ad66832e3cb28a03dbad7ed650ab43,e8ea449e622f01116eb55995fedff35f2fa052bbb6365a90accc9029b7fdc0fc,8d0311088fc5cfb8bdeb11e0132152946e19626dc01e66af42f6fb613a220a97,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,31.313,31.3,exact,canonical-source,profile_cf49abb6dc20240b5e194172,reported,unknown,,unknown,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,sre-bench,SRE Bench,terminal-agent,4e7fa7412a8038632d762711da90f627da26278faec9ac2add3078ffd04134b3,1ce16da4a13f5017153d0fb9580ef23f500b602218669d9cb734c58951dc8754,5932620a3562e6c15642c8cd0631f76dc739a777aee0cd6f8a3a473b4f5db66e,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,3.053,3.1,exact,canonical-source,profile_ddb840c99baf1533c559c1ff,reported,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,apex-agents,APEX-Agents,tool-use,f31f22c684091971402ba31bd9d4a0b19ab4347490541c5e8b5600d492df96f6,267297a74c1db7d07e589fada82c1eec492ba30767387c0d733a6c83e96fcfd9,a467ee58ec0fb95e7179d89e476f72b01cd22c60b3880acf04ff963bff40cef1,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,54.400000000000006,54.4,exact,canonical-source,profile_3d1d8a68b3013e79d3d50a58,reported,unknown,,unknown,,false,2026-10-19T22:24:14Z,fresh,12,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,arena-agent,LMArena Agent Arena,multi-turn-agent,ef434dcdbf7d9a19ec238641f8092b607a9b029296324b8978d75fe71dfe2750,34b097b5f9b54228a8d53780f73a7bc7fcc56ef78c39b794a3a152cef7b628d6,08cc136206cb6cb8a231ee320a7bdc747e451bc1013f7a33ccdb1b2a5626b343,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.027705996772957868,0.0277,exact,source-specific,profile_fa699564a69ec0aaebf7c223,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,2b5a2b7e76f9d6a28feee5762bd6007c7253920ea380e8459df6b6f628e98967,760f3e57e4ab04d90cabcd308237af9162251cee2e15f52bcf64d61582a12525,9fc9b7c83607a970df00b9e96439369804698984fe008f5d709cdbeb984709a9,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1689.4,1689,exact,canonical-source,profile_ad5c4858e74f428a0183b9f8,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
xiaomi/mimo-v2-6-pro,MiMo V2.6 Pro,Xiaomi,open,60,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,2324bbf56bf5f5ed764c8d52ad3bb18defaa2ac03abbe1a68150d60b2cb4844d,68a1426a1962b935654dc641b44caa05687b5db853530d156c990836546aff80,986b2f8ce1ae7e2e588355a186edbae80339671019098dd79eb65d244ca707ec,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,2.857,2.9,exact,canonical-source,profile_456c16d235a741750dce26b6,reported,unknown,,unknown,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,epoch-eci,Epoch Capabilities Index,capability-index,19836fdd51065520874ac5af8481ba861c1462c41e7bd2b7b69c3db4585a154d,e03ef36a48c2b6fd3e17278c201d9c332d2d7b45c419f07eeaee5baeed63de6e,74266ae747b3a0ae4e58b29d996911803a428a9206c19520222fd35081db9efa,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,146.92,146.92,exact,canonical-source,profile_a112784282b9946d07ab8413,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,3c52bce955573f887851e9b72cb2a5a2e651336fa8e07c2ae07d3025344c7610,03f3992a23a8baf2a82b49b6fef371f93ae0061dd4404ff523c541f8adfa8645,94615ddece2f787acdf0a8cf1e4adca2824db332a896625bd695b6c8497dff37,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,18.728908,18.7,exact,canonical-source,profile_18bf23dd51a56ac5e133a7c1,reported,unknown,,unknown,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,arena-text,Arena Text,preference,71a16bb10468300259049c51fd615f9a1cbd727be37b79d5e458f386336bea42,c8f535b37f5f30c3deb24d1217e50645a5205c67c092859c0139bd6becffa0bc,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1439.2396535610842,1439,exact,canonical-source,profile_ab1b92201926a7b0c5c81541,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,livebench,LiveBench,general,60fb50acd5b1359541e0caa383e86da6128c9dcca2c4562328917426e814e01c,928c7fe84c4b07770e556c22703d6674690d1ae9de1c3de82d5907eb8db4cc98,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,67.25676190476189,67.26,exact,canonical-source,profile_3197178f18af61acd90327e5,reported,unknown,,unknown,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,2923ad267069cf57704fb4d7d106248a37a4e681f2b6c335b0ccb204bcf32199,fc2fed03130b8fc7f91310105a18424d36ed8f8a539f338f7ffe17125c8c2563,80ca7e79dd41fa96e23232818dd5902735c42906292664f7177edaf240580de7,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,38.97127,39.0,exact,canonical-source,profile_18bf23dd51a56ac5e133a7c1,reported,unknown,,unknown,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,gpqa-diamond,GPQA Diamond,science,2c356de3dd532c1c14b951b713e231b204910cbf082a24e7212927d3944f8669,1a921d2a9386ae169630fd197fa644712a1dc3a1ecbedbb26c6c953f15e2588a,1b15d9cd60eee77accada1992be5cbb948f8802854924763ee21b54e02ddc26f,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,92.676,92.68,exact,canonical-source,profile_6a5ac3a55dc02af104cdb177,reported,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,mmlu-pro,MMLU-Pro,knowledge,b68626b38cb39712c215197d1dba8fe21485054967ad06dd67966f5a19f4dc9c,a9b67669dd69c010c1b37df413981759e49768c6356634d53f0ec6124dd93494,68c3183d6b5d2374a63aa8392e3467e90f39614ebd12ac078518b4263bf7cbb8,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,84.22,84.2,exact,canonical-source,profile_b30a7dc7d7d151c591b300b0,reported,unknown,,unknown,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,2d1879e2b2a4710c88a90963c4af6bba1efeb3267193cfadce3330e15508f96c,dd2463cda6a0b929188064149fa18d987669ba54708a160ad626f32b461acbbd,217614212a82bf56f939040ce2555be0306ce9ff38db111469c28fc88105ccf7,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,26.666666666666668,26.7,exact,canonical-source,profile_e15386937757ec3e2088b88f,off,off,false,none,,false,2026-08-10T17:46:50.000Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,mmmu-pro,MMMU-Pro,multimodal,9bc442fffb8cd80a576775afcc67b250c4e25ff6a64f6089862cc2d54d03a490,a633804e6f197472abb1bfc49fa5af38c48ec3c9e70855f2e25176c61e41bf90,9c002f8540b8a0f6922c6c41e68c2963a44e94eab38d84e8b5bbd4abb56a3dc2,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,81.156,81.2,exact,canonical-source,profile_d535e0c60ed655f9066eb7da,reported,unknown,,unknown,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,critpt,CritPt,science,26ebfc4a0293e189cb421c54edc42a220dfacd547ba8af7a8fe74050c33e84e0,61cf861e03d3e4312e23ef492721a287b1513d7ef31c00d252fe09c4a0dd6af2,acd39e2ee82d0f5134f524acd25abcad1cb1d2e4980790f737057dc3f6759f54,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,3.714286,3.7,exact,canonical-source,profile_18bf23dd51a56ac5e133a7c1,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,simplebench,SimpleBench,reasoning,47175f509793812eb12100bf1e54e8ee9f1209f96e63635bf8a060f53d3d7203,ac4d6aaaba5e665f4ea3e9803ae8d66d8f422de0dfa3919641803f8e74797384,675ec8382f52f65e03381904418d26e916a0f19bc9b8754d277d20c026700fc7,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,45.8,45.8,exact,canonical-source,profile_8621ac6c3391417ec3519f00,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,chess-puzzles,Chess Puzzles,games,2303aa1ff0dd43ee4b1fdb30e4fa5564410c9befc0d489f184f3c118f95418ed,817f69e16a1d5d62146b719e6e4eb080f0ef1f82123d9be930c6846747f558cf,074464460b8ed4ba876fa15519cf3ef09d6b8fd3c599c005f425e28ec58a3145,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,7.000000000000001,7.0,exact,canonical-source,profile_e15386937757ec3e2088b88f,off,off,false,none,,false,2026-10-13T13:44:41.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,swe-bench-verified,SWE-bench Verified,coding-agent,ffd246b1b2075ceae84c43285a7d423faf621c6f76dc7719748a22bda5eabee4,47c4b19dbea4b1d07d736c78fe6f247f5f919abc6d904361a9847b7df4589a4f,6d2ec613055dda82bbdb11ee8ff410a0e5080632eaa5d39de43148dbef744585,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,75.0,75.0,exact,canonical-source,profile_02c909f667465bb33109fcad,reported,unknown,,unknown,,false,2026-06-02,fresh,152,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,swe-rebench-v2,SWE-rebench v2,coding-agent,8945dc6dddb53f0ab9024730c0cd81936de3b1c5ba4b3c04f660ed32c8e26ba4,0a4e559414c29b68877d607a5553e35ebd2551fc5ad5f53dd3ba808cdfc143a6,9b03de353448bec4c09d54b83315b7951e30c2d8dba467c278068f98a4886798,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,47.207207207207205,47.21,contamination-warning,contamination-risk,profile_da68402d6b49d63678a6cd3c,reported,unknown,,unknown,,false,2026-07-01T00:00:00Z,fresh,123,0,A+,0,A+,7,C,potential-contamination,2,potential-contamination,swe-rebench,SWE-rebench,https://swe-rebench.com/
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,frontiercode,FrontierCode 1.1 · Main,coding-agent,fe4e1b52f686213cacf8642bbd44f73431baf83d3ec95982458fdc354aaf3523,befb1c49ab8de3856e086add7cd7937f20c7841a05ae81ef2d62b2a48bfd1bf8,22dbfe3fc083c8b5919e05121ea8327563fbfacbe07253034fcbf3f801d471a2,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,14.7,14.7,exact,canonical-source,profile_e15386937757ec3e2088b88f,off,off,false,none,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,livecodebench,LiveCodeBench,coding,bb19223f206e3e728b21e8cf2187afdab2f1572fafb40afa00498bf7f5a10bee,23c09023a043c2c8adb9480b0c1ecba4f6b21ec128531b80038e266a73893abd,b841660719bf18b04a6287c32f9cb5f406460b672b26c7f614d6c01c7c546f22,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,82.154,82.2,exact,canonical-source,profile_ef51fefcba684ed7611f3d0c,reported,unknown,,unknown,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,scicode,SciCode,coding,ff692fb398b5d2fdb89d52d5153a63ad9e97fd41b0edfb26e230f62faa665a90,2749d7e26d79405110b9cafff53ab2531044cfffdd5edb2608c515b514d7a19a,ef33c85a77b01f2b6c6767d505eb57e69b4cbf6c39ccdce61e2de8153203dbb2,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,47.1064814814815,47.1,exact,canonical-source,profile_09b0d8ecd750138e32e398a3,reported,unknown,,unknown,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,ale-bench,ALE-Bench,coding,67c15862877cc04e1b5cbe8acdda3a9c352f7cc04c68a2edd1355a92f74c0f1b,d26a00ca5590786437bfdf24453df124e479a92f836fe8a85a9d44ef494d95dc,8b1229a65e9d84d7833714acd10b9efea2d02da2e4034328743cfdc9dbfc1d64,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,640.02,640.02,exact,canonical-source,profile_bb51e947973ce8b8932407ec,reported,unknown,,unknown,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,vibe-code-bench,Vibe Code Bench,coding-agent,c483088299478f2fc65c3562fb7798d8e6e549443ddd7e42482ca046770f9620,c374f2561b35c98dbc219cc222cad2548ad866d912b5b3073ce3bb8b8bfab8f6,0a74c5185d94396ea711a709d4e9636711650b9d504adb512aa8d35508c2c8bb,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,47.566,47.6,exact,canonical-source,profile_20bf9dd5ef508261c42af9c9,reported,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,webdev-arena,Text Arena · Coding,coding,015e9a5fba37a42436f3fa70d73271219ee58690734a0599f2c16197177bde42,5ccbbdc00afd2e219d80993d141e2dc2d2034a6576ae92c0dd91a13234a2816b,2402163598f231bc3d9df99d4df7b769d77309c8c50f1aa56f9096b9305b2071,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1483.3051511403955,1483,exact,canonical-source,profile_565e2d3f37afb7a4e6e7b33b,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,6354f79e98a147a9255d340d725ce2058fc7908465d7e76841c53fcf0a19270a,112f8417e3ee4d113b7629830f152fb1c5635348977b030ed570f2bb7edca495,0fa2a10e03c0b94d9d80bf37b1f554b84b05ef7c9ed5c835ee5f901016d1589c,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,53.558,53.6,exact,canonical-source,profile_abbe9505531d90fc1ec94ed9,reported,unknown,,unknown,,false,2026-06-02,fresh,152,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,2c13dba95b385f5f61d3ee73b77e7c6cd105295c696b2fa1268b65bb52c66afc,ec78dd4bb452f43f8bda80243799589872a7276a6e0595d15000cb6fc5e577bb,b1a5a7559b54b8b256544d9f1317a243cc3e3792ce9dc6ca4c08b3bbce7f8b82,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,1.01,1.0,exact,canonical-source,profile_c48513ab822d22c8f8ac8b40,reported,unknown,,unknown,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,6dec54c4d5cc12f5853aa4f8df893dac34b49887284106f09b93fd9552187c7b,3fda36d687a9b98c128b39d065d2523034df8317c52dd03f33afe8b880c76e37,44e9ebc0dbd6a44ba532a659443b94234517f1ab1c44aa3f7b045631243bed0c,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,15.257732,15.3,exact,canonical-source,profile_18bf23dd51a56ac5e133a7c1,reported,unknown,,unknown,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,mcp-atlas,MCP-Atlas · Scale,tool-use,35c45af3cf8f52f9ebb0da9f18dbf2c64a2c55bedbd0089f45af09c14e12b643,9ea4c8ec00c3440e32a090e84d5a316630d9aa766dd6b083f2938641a9b6f7ff,e053c3facbfcfe53f448c08c4ad0fe3a017da18caa4fcf1bf742d31d3777a4df,release_4873e5f09254ceba3d71ac2e,MCP-Atlas Public Set,provider-official-code-public-set,accuracy,percent,true,74.2,74.2,other-supplemental,supplemental-source,profile_eb4c46e38e87a4548c65dc96,reported,unknown,,unknown,,false,2026-06-01,fresh,153,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,minimax-m3-report,MiniMax · M3 launch report,https://www.minimax.io/blog/minimax-m3
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,skillsbench,SkillsBench v1.1,tool-use,42d0b77222d0d07636ff0449edee1b0ecf9181377d55db26755faa11f290bd09,a00f5a464fa7216792c4963634252e872ba0655a026193ceb678b311a2241752,c2ed2025e534826e5e627446f4493a128269710491b35eb2122abbaf19d56304,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,53.0,53.0,exact,canonical-source,profile_09be712f9836bbf7d088767a,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,apex-agents,APEX-Agents,tool-use,303fd0f4d930e6a70da71875a708a3c9039d231c30e37451ac0e0830bb3a498f,e13272da54a4d290673b0ea40a7a0301d89d171a187ae71d42fce271e1142936,d47ccfdf0d6ac215be0ca50fcd3c33ad4faefaa3c8de942ecc63915ed8ffd39e,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,37.7,37.7,exact,canonical-source,profile_53416c63888b302d0224ceb0,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,arena-agent,LMArena Agent Arena,multi-turn-agent,018e01bc73fc3787b08f7601fd500872a6d36067abaf3f173bbda841c96dd8a3,8e76dfc0a51a83e98f6f9555077b683c81debc5c1cb7fcdee0bc0abd233d8676,3f170e27244e2f1ca1729f0584582ba73dee3add65a2555309ff41f8c60edc56,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.10161684431257027,'-0.1016,exact,source-specific,profile_b5748244f086cfcc6a37e5d2,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,vending-bench-2,Vending-Bench 2,multi-turn-agent,cf75b59424a33c89f6ab48ee0841f64a445af78d74ba2eada08d3ce69f4f78b4,bbdedb5475b983b3b835232476ac4506b7388da4d1abcd6c5c17d9885a0ff086,de681cc903cad707fbf63a3cf8ceb5e6ebb6489f2344af45c1ea84c947480979,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,2157.7716666666665,2157.77,exact,canonical-source,profile_951eac4e039a4dc0d2cf0bf1,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,ed6fe6a45d533f61175e49891805a66a3466e4846e81213ff5d3dd8ec065aaad,c40203704d883048d9cf584caf0c98e08246b17b727d908874c22cb9fd24e46c,7965dff76749559df20e3509d2bf6e54f959c520febe774b2aa356e982abc5a2,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1249.95,1250,exact,canonical-source,profile_18bf23dd51a56ac5e133a7c1,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
minimax/minimax-m3,MiniMax M3,MiniMax,unknown,61,osworld-2,OSWorld 2.0,computer-use-agent,93a2a73d52faa1fcd519ccac761dc837cac95adddf497e738584ecc943322cd7,05192f1dc3fcd0ecb7159fdb9f4ed6e425319513a4f247cc61929de333ba8e5e,fa4b66d5a1199a067fe5b1925d4b19605dfcddffa329b61b71c74d6d20670efc,release_15c5525fe39f33bc181dfb69,OSWorld 2.0 current leaderboard,binary-accuracy-500-step-budget,binary_accuracy,percent,true,4.6,4.6,exact,canonical-source,profile_91769ced1b80cf6db8086b91,reported,unknown,,unknown,,false,2026-09-02T09:27:06Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-osworld-2,Epoch AI mirror · OSWorld 2.0,https://epoch.ai/benchmarks/osworld-2
meta/muse-wonder-1-0,Muse wonder 1.0,Meta,unknown,62,livebench,LiveBench,general,cb7f4c13f93776ae59fa3ea1677554414467db5bbfe8424628a49807ef84518f,309069d2c1db92bb0dcd099f160618e75970d3f7daee444913eb2282a82831a7,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,81.15589285714285,81.16,exact,canonical-source,profile_c64279d72fe7e5e71cca36ec,max,max,true,fixed_effort,,false,2026-10-13T00:37:05Z,fresh,19,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
meta/muse-wonder-1-0,Muse wonder 1.0,Meta,unknown,62,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,94de55866ae5b7dc2b9ece154021e0812c92cc8d21e02ba03e7a5c0427464495,25fc4b7f006372b4879363f1c0a2ca97ab4629bc65e7a9d62256581af3d425df,c5e8afe0b262bb13161157a37bf4215e1df3d12bb5c25ca08444f51bc477b669,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,99.44444444444444,99.4,exact,canonical-source,profile_c64279d72fe7e5e71cca36ec,max,max,true,fixed_effort,,false,2026-10-12T19:00:20.000Z,fresh,19,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
meta/muse-wonder-1-0,Muse wonder 1.0,Meta,unknown,62,frontiermath,FrontierMath T1–3,mathematics,a038a6f431f098286f786edea5ff1c0282483f84856464422a834138ab641649,e5f4afed9fe2bbee92e5d80c25416f674f31d345440dda576ca9051758dca4f4,865164b37968dc801d2248cd52aae4434263b2b3c813983ccf41c155b2d69f32,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,82.10526315789474,82.1,exact,canonical-source,profile_c64279d72fe7e5e71cca36ec,max,max,true,fixed_effort,,false,2026-10-12T19:00:20.000Z,fresh,19,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
meta/muse-wonder-1-0,Muse wonder 1.0,Meta,unknown,62,critpt,CritPt,science,e67d15ca59300b3c1fcc6e405d58bf8c53cf05116c0a3f14fae35cba534de8ce,851228a218338a012c66f1babf8a779c06736556cc9fb146d6f98ecfa4f72e54,2fdc53687bd0c3c9e45c0d8be86629a4284fc54de5a01963bf2ca2fe71c844e0,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,28.2857142857143,28.3,exact,canonical-source,profile_c64279d72fe7e5e71cca36ec,max,max,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
meta/muse-wonder-1-0,Muse wonder 1.0,Meta,unknown,62,swe-bench-pro-v2,SWE-bench Pro V2 · Scale,coding-agent,973164c7fb636ab3702cbef35cc36491a1950e3b25d20a9cfca26018e28bc736,b8c6c437841c15b48f6b40ab5696dab4d8f0a9a598bc0fce0bcac32a5611aa45,846a8b0046459da9491dc2ef8a2af518603e61eaba22045fbcdf969c35f90b32,release_d9b58bd286ce316ccf0ed891,Scale SWE-Bench Pro V2 public leaderboard,public-v2-642,resolved_rate,percent,true,96.73,96.7,exact,canonical-source,profile_7ca4850f80ce42907607f3af,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T17:57:23Z,fresh,17,0,A+,0,A+,1,A,independent-evaluation,3,independent-evaluation,scale-swe-bench-pro-v2,Scale AI Labs · SWE-bench Pro V2,https://labs.scale.com/leaderboard/swe_bench_pro_public_v2
meta/muse-wonder-1-0,Muse wonder 1.0,Meta,unknown,62,cursorbench,CursorBench,coding-agent,fb09f5cbd4667218b8d7dd2c2a76d6ddfa1acdb76f46d8c337aff9c8981ba2fd,35d58bf5ec71b6be183254cb4e4cdf872031da6a9567f0797c0d0bff8e90ee0c,1a4b83a2afdac24df41d745c36ec531d1c186b0988262fb09cd7221a7227996b,release_9acebd5b002e9c3f18a30606,CursorBench 4.0,official-current-suite,score,percent,true,50.4,50.4,exact,canonical-source,profile_c64279d72fe7e5e71cca36ec,max,max,true,fixed_effort,,false,2026-10-19T22:24:13Z,fresh,12,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-cursorbench,Epoch AI mirror · CursorBench,https://epoch.ai/benchmarks/cursorbench
meta/muse-wonder-1-0,Muse wonder 1.0,Meta,unknown,62,scicode,SciCode,coding,ca1849dba515473640c165bd0a9a5eb41a7e44f00825c2c4865eadc53e0255ea,be4ac6bae072f03a2c0dbf832e3459391dbafdfa32456ebd6da82f62ba75a74d,e7f0d559b2832e84ec054a7cc453b0de14d649cb9c59eb26f4c80a1541a67067,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,60.300925925925895,60.3,exact,canonical-source,profile_c64279d72fe7e5e71cca36ec,max,max,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
meta/muse-wonder-1-0,Muse wonder 1.0,Meta,unknown,62,mirrorcode,MirrorCode,coding-agent,b890cfb76fd11a0b9088fd52def87f22bb6d538908342594b9b6f0b79d02b9ca,11820b02a1bac776eb351015e6433e3bea83aec8323fe8d5f7a4b3eadcfdd464,f33a4a2ed881566d0ec3a163b5c816bc31b172e3871656ddba3fc6d53797cf84,release_07bf24088c0805f7b6df7a88,ML +Private 2L,15-programs-30-language-tasks,mean_score,percent,true,45.55555555555556,45.6,exact,canonical-source,profile_c64279d72fe7e5e71cca36ec,max,max,true,fixed_effort,,false,2026-10-14T15:05:06.000Z,fresh,17,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-mirrorcode,Epoch AI · MirrorCode,https://epoch.ai/benchmarks/mirrorcode
meta/muse-wonder-1-0,Muse wonder 1.0,Meta,unknown,62,weirdml-v3,WeirdML v3,coding-agent,25e598bd6ef6413d291228270fc74fee6ba0bbef5169f55320fe1e6569bc68a6,336b5cb43a649834c0c1862868867432141b38e1004603402c4e0b98ef806dd3,867c92cbe6523d17c2506e8f1938872625ec5c7095976601a846815178e97ff9,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,15.191042532467533,15.2,exact,canonical-source,profile_7ca4850f80ce42907607f3af,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
meta/muse-wonder-1-0,Muse wonder 1.0,Meta,unknown,62,apex-agents,APEX-Agents,tool-use,cbe04502368e748138f2830696bf1f5f7aa79589b66df70cb18cd67fa2018836,9ba1605e6c3feb4576996a6cd87f208a17a38949430dd4e0fbe6dce884fc4e2b,12016967d2e7faa9938f5f386ee5de7cce5e4efbe5099b2df6614195fc385102,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,66.1,66.1,exact,canonical-source,profile_c64279d72fe7e5e71cca36ec,max,max,true,fixed_effort,,false,2026-10-28T21:25:02Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
meta/muse-wonder-1-0,Muse wonder 1.0,Meta,unknown,62,vending-bench-2,Vending-Bench 2,multi-turn-agent,2cf48b22073fed44be09cabe5737687d9dec02479b8d4d492b0fe4346d65ee83,05e58c81498a7587ae45a81bd0dde7cd7f2d01b4e3b648272fcaef1cc5596ea2,ca4b31ee27bc9d049375d6303d49804bc2d7779ba209152d9a7ed9b280e69cf0,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,9432.18,9432.18,exact,canonical-source,profile_c79a10b1e857fa0c53d86a8c,reported,unknown,,unknown,,false,2026-10-14T07:48:15Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,epoch-eci,Epoch Capabilities Index,capability-index,a485e83ca00b39b2e4bc21c25cfcc478a35edb61c78526b8d135d2a88bc2ba37,f6c2451bb18913c1816ee588772c75c70b87fb899c422f026d81ea2749f4c0a8,0b21fc9f1947236f2b062e2b2f0ae739db3353a23d4d3dbc8e9c4eedd81c6b18,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,146.2,146.20,exact,canonical-source,profile_eebba79f988efd30ca30bc67,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,arena-text,Arena Text,preference,85e242010bd5a9cf7fe64a24d557be68d5e680ec5523f7e3a8fe8f49bf2f3db9,64774c3307c4f152815bb478820a530bfc69911d72c0d5779cfcbd890e0090d5,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1438.6546684400676,1439,exact,canonical-source,profile_ec6f9dd70348c797be90f74a,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,livebench,LiveBench,general,93a93d5d5ea0e0fe97f3281bc14a678c041776ae10dc5f238d4e69bdafe7f4d4,5266991625c4f00094c13464dd33cedc6ffd88cd1ebe70395462033e92424db8,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,65.48110714285716,65.48,exact,canonical-source,profile_4205393e3ffc155c0364a879,reported,unknown,,unknown,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,7ea3e8bc80cb124094bac0118ba8cd4f238750b1cff0bcd9f531e6c85cf2cde2,4e2ebaaa6dae65fda59f90543a4b050e62959f217320f102bd15cc717bda481e,9d79f00315c23bc47d2293c9003ea42de6ca857854c04113ff97996ca36b3878,release_1f4bfa5486eb4deec4942449,HLE original,provider-report-no-tools,accuracy,percent,true,34.8,34.8,other-supplemental,supplemental-source,profile_a5bf84569eb675351feaaf13,reported,unknown,,unknown,,false,2026-04-24,stale,191,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,deepseek-v4-pro-card,DeepSeek · V4 Pro 0813 Model Card,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,9c86df606b22821cf0a663e2b9e45f0e4845619858e98d55cf1b3e3c154aa2ec,e06c4d8352f94395bb4a55ac011df146b2f51f389a831f127f434bf1ce7b17dd,79335436b03dfb862b6194cf674507844905c3d45289028b9afdf4387cc3f45d,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,7.784986,7.8,exact,canonical-source,profile_eda89840d177ea9f880a9299,off,off,false,none,,false,2026-09-09T23:36:58Z,fresh,52,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,gpqa-diamond,GPQA Diamond,science,0648c2c0a94248c004169311bcbea5cbbade5805d062a1ecc6d748e48a6e5518,c48da74884a27ae1895b897d5a74bdac3407a5a6f03cf389a5d41977e6eb3c00,ffe952529fe1690559e859136c0741080672196f94432543cdc2b55e58ada696,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,71.616162,71.6,other-supplemental,supplemental-source,profile_eda89840d177ea9f880a9299,off,off,false,none,,false,2026-09-09T23:36:58Z,fresh,52,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,critpt,CritPt,science,6fa586d006dcc29aec311b32793ee785e08f5dd1d9d24cf77a5da287ba4ebd69,29e5d7206e0d2ed52ec89713a13e29db658fbaf682f4940faff5e927361dcd83,3ac97c3206650b74f25224218dab2ddb276c85687e3e1ab6c72733caf61af50f,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,0.285714,0.3,exact,canonical-source,profile_eda89840d177ea9f880a9299,off,off,false,none,,false,2026-09-09T23:36:58Z,fresh,52,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,simplebench,SimpleBench,reasoning,1ac33e3ed1f409da3e5af9fff3e1245aa9824110f9ef0ff0f6caca6ef77f4047,4d8c5f3cc0c3f0e1a43ff38d8239c68583e59aeb581768d971916e7e3bb73888,2221c745b377caaed9b0d2163c8c860d834304111b9d41acb2983d63b8dd702f,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,46.3,46.3,exact,canonical-source,profile_1f7d663cf31b9ff28820da09,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,swe-rebench-v2,SWE-rebench v2,coding-agent,d4102e0fdbef1329b958da4b9a2ad911e0afc0f96e4e194b1c76a4233a5720d2,3844059c64b7ee0430d0b146b0378a0e4ae15763234634cbff5ad9b5e0962505,3ddda07372b15613c728aba016bbc85e044ac6f484d9590196b5cd0f7c102c71,release_332cbf0e979daa9e173de9f4,rolling-v2,2026-05-01/2026-05-15,resolved_rate,percent,true,40.0,40.00,other-supplemental,supplemental-post-release-window,profile_ec6f9dd70348c797be90f74a,high,high,true,fixed_effort,,false,2026-05-15T00:00:00Z,fresh,170,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,deepswe,DeepSWE,coding-agent,9238619bc18e226d74a594c59bb2e9ffffbce91fe93441b3a9b4e2e2d62f6049,af904b09b8ddd41cddb0dee730909017d6160fc6fbcbb23b39d6d457da301045,e858d0bfc6745368cc291d05a20f98f384295dfe782269142355d5d72adf3db8,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,53.31858407079646,53.3,exact,canonical-source,profile_277305b76ab123eb94f5d2b8,max,max,true,fixed_effort,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,scicode,SciCode,coding,637343f95289d46c0579a54f45854945c32dbcb483c02e42b1a6a9630a78f045,6dcbc9e25f002ab22ce73feba69e87f6ad77ddd8f6ebbd9b71a6d21d30942a7e,1b7162bf1583ff5201fcc2c31296c7dc34046178f2bc03b1dd3fd3a5f4c101d8,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,44.907407407407405,44.9,exact,canonical-source,profile_277305b76ab123eb94f5d2b8,max,max,true,fixed_effort,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,ale-bench,ALE-Bench,coding,b89e55d7d114ab8a741245542b20152bd6a24fb27e2d8f6eb797221a73a4a3bf,3f032ba9fb6f3fc48ffb803cb6059d8c00cbb5b4ee395a41dbd8496dc9a77e14,04e6ac1332540961c0b5a663bb13fe5851a752dec7ac4d9a7e4c507ff8094681,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,678.2,678.20,exact,canonical-source,profile_ec6f9dd70348c797be90f74a,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,webdev-arena,Text Arena · Coding,coding,9ec17120eed9fb2e7fb3350adc03be5395d03fdebed4398b12b3fdab7637535c,ee4026ac4006cfb4fd836d7fd2220369bdc2b86a8573e71772d468f0ac30a3d0,67f2d4b9fc6ab4c17eba26475c1a52ebd4b93b2355d8fa5e2d1968900cb10256,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1430.4821164472921,1430,exact,canonical-source,profile_ec6f9dd70348c797be90f74a,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,ca4f42e7973a5ae25c4a66240134a0079e353cd3d069bba98423059c5085ce4d,d8e97704de6d2539f0af3ce07dfa4fa815d05fdbabae8e24d0289116be968379,2856f01ce5091a1b0fbf9afd2ac976cbff4761be067a817d79b63afd325fbd4b,release_83e2a53bddae6e68d7ed56ed,2.1,provider-reported-terminal-bench-2.1,mean_reward,percent,true,61.8,61.8,other-supplemental,supplemental-source,profile_a5bf84569eb675351feaaf13,reported,unknown,,unknown,,false,2026-07-31,fresh,93,2,B,2,B,4,C,secondary-re-report,4,secondary-re-report,deepseek-v4-pro-card,DeepSeek · V4 Pro 0813 Model Card,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,skillsbench,SkillsBench v1.1,tool-use,0c6f6dcfecc265f7b63ae4e140d08e2d49cb61563b28f532dd9d760fb8f7e648,9a1d3bb48584137ac53d1dd9f027f9f5bc3eaf08fbd57ba1e1ae72f4335950cf,c2ed2025e534826e5e627446f4493a128269710491b35eb2122abbaf19d56304,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,44.7,44.7,exact,canonical-source,profile_4ce18cb91b2b4b478d573def,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
deepseek/deepseek-v4-flash,DeepSeek V4 Flash (2026-04-22),DeepSeek,open,63,weirdml,WeirdML,coding,701753a8e1689ab6d9ef13c7630123b6b83409de62e8e749c117ec64c3fcbf31,cb5be2c791183232a385105a380493966a2ce01edc3034759a2aaf99f66304e9,c7df614619f26d191e0477590f0757699f3397570d859494fc0fdbb2c29afa98,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,45.63,45.6,exact,canonical-source,profile_277305b76ab123eb94f5d2b8,max,max,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,epoch-eci,Epoch Capabilities Index,capability-index,ede14bd53c66239b260ad547a3ffb2caa502d9e7ccec339afda20de43dcbe05f,a559f724d779d95e1f6bd9f0331c45852172df7a76288a7e3832af7745d36048,ae6c569248bf0da4a0acd1aeaf38b0a976f14bba98a396b489c822fcb8b1b25b,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,148.82,148.82,exact,canonical-source,profile_0c4947d15e8e8b2e1d736e54,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,arena-text,Arena Text,preference,1bb5cd55dbc951ac63a3103307516811107e4fe42d15f562da1fc4b528c28992,ed2346510b3e5d33b4d82517046d62e7add5b91c063eb02d811e0c0878c965cb,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1447.1749944782787,1447,exact,canonical-source,profile_cd463b62537d29712133caab,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,livebench,LiveBench,general,c542e5df54838e4a4728b3f017a3c1aff079bb28fb8cf9eb2f991f714f1c0a7b,0ded6574c296b79fb1ec97c5212bb713302dccea22580b5f6e65b823ba3e880c,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,66.37396428571428,66.37,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,simpleqa-verified,SimpleQA Verified,factuality,e21f2c2885d5b75b6821d0eacfb7c3f03756db59ac07d087a6a6542762a43fec,ceb86e8cab7e425096679c47d9cc18235bb62f8caf5958bb908228dd1331f406,72fa00adc85d3388e9d9fc1c13acf750aeed030934a81665e85e3976b5089d87,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,29.4,29.4,exact,canonical-source,profile_cd463b62537d29712133caab,high,high,true,fixed_effort,,false,2026-08-27T19:28:49.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,gpqa-diamond,GPQA Diamond,science,dedc5e6df3fcb08813b0d740ded070aadd42e902fc85c7a224c418b271e47243,f4bf0023f19ff09fd23afa1813785d26f62e3c7a97cb6c4508cfe694a76371a5,bb5ed5cb8f0d8db4cf55d5128bc1072501d648d4acc49ee6d831310a9bc919a3,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,83.08,83.08,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,mmlu-pro,MMLU-Pro,knowledge,5329cdba31001e5a0fe134f9465b70b2aaef5cdc49de82400757f42713fab78a,43b8e3ef40ed531d8826d189b2cbd0ca3f465a1cddd7ef7fbbd1ed8877958df4,6f17099ff66a523d4d6a6f88b32a39f321eb1acd5e0fced00386a94872e580af,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,84.554,84.6,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,9bbc062b4fceff66f9a7d5d415e859138477092883a01492f3c6624abb01ddbe,f11546dabf32df439dd5ad5d7ca1b54d4d878aafb8c55394d560c70dc939b149,46c32a6dbd5cc35e27c1a07fcce0749d6df3838e83710dce955b3c864b4e92ea,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,88.88888888888889,88.9,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-08-07T00:56:59.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,frontiermath,FrontierMath T1–3,mathematics,c389e27ba1e1dcd89349ce735371ee826b5d2e65b467693027348a6845611c28,dc04a8df58946cb7f965389054e3cbca36c821ed3c3060dddb7cbd7f8f06a740,c29be5d216a0fb49320061c3ebdf535bfc5fd74d6084a85230e48503aa015a3b,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,51.2280701754386,51.2,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-06-12T16:40:38.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,arc-agi-2,ARC-AGI-2,abstraction,7a05f4cfc25c7876af6e3f4e2960f4cce2985efe9acddedeb6eeb574242a6bbe,bb0eef2b77b1d76edd7e612c8c94fbbb6b47baffe44d63034953cde75bd57dbb,e80a40dce6d44e9548d8b5cf7f58dbd5b98a2218d63a513b4a980ca9d1acb61a,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,18.9,19,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,frontiermath-tier-4,FrontierMath T4,mathematics,c340d38e1c989eec5795a20d4374a4ede9f40e8d410c0f64b94ff2d49be56cdb,7aa5032e387cc68152f8354e9efe25a10183d917b129ed6b7a87411647368bc1,3a6f043e3447e4f181bc0555c74923aaa94035d15034c142b42db7f783a752ba,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,9.75609756097561,9.8,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-06-12T16:40:39.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,mmmu-pro,MMMU-Pro,multimodal,964e853cdd98fe992ddef66dbcb6e69a3313e399e021523febd69fddd0ba0711,91dd7a1926794c16ea6c27dea47b23da459d801020b36cdf5e1e34c5a93eaec2,60d5afae34318f4d140d29bfe7d6d12e78996eb082871bb2ce95e5449f760843,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,79.249,79.2,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,critpt,CritPt,science,2b86689e9493efb86d56157a45d18f182782349032edbb8802083675892c7853,f47350749ea3dd392f3a57c52efe326854c1a1d7dba9680ccab8715d481beb68,b9c8600f4fb5911c6e6a2c59f7ba2e7e43eacd49781d70c54cd9faf155510642,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,10.0,10.0,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,chess-puzzles,Chess Puzzles,games,1f2e1386bcb4f941d96e8f125796d75d47999452e3a781b14ee19411000e79ee,2e793651f82281230b768fffe4248a8a2f93aa87370b255f786a27be95767c17,fbe42d1459879fe0087939401722acd4127a2251454373d72271b27aa2489d1e,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,17.0,17.0,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-10-13T14:02:26.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,arc-agi-1,ARC-AGI-1,abstraction,2281f723afdbb928874921ddeb071728a9c5bb1a7a2d004231dc99ed5e3c9052,8e3a6ce445cda7ae6288963025851e41539b91a1e5e280cf4383b588ae479730,ae6f6765b96f58fc9143382fe604f5c2fa58622e65ef712f44f33e10871bf936,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,63.67,63.7,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,swe-bench-verified,SWE-bench Verified,coding-agent,b5eb5a131bc1544c212e3bb167b57dbe0aad4baf8e754055b679d941f4d14703,1235b6fd91e45720eb84724650680b0f6fb3630a17997fe4b93542b80038cbf8,deddb0dfbdb6baed3a07d79ef533ab29aaf4b2085dbaa218755e23f2162181c0,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,73.0,73.0,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,frontiercode,FrontierCode 1.1 · Main,coding-agent,acd2b5e325e17882a1b1ca155245b0315f891aaead64d37af27d6c30f3984a01,ab11ecdc6ec69cff9e0534623a597d634d4ad63c77592cebc2bba02b8011cee6,a914ef90457ec38e014bd6d0c4fa13247b98e205edefeaef436af0da0b0713f1,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,27.04,27.0,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,livecodebench,LiveCodeBench,coding,f83374d904ae3ac3da645629fd9087f42b1e22ed671ef7ca193e846d82d7b28c,38ce3e9a9bfd785723b7c7c9a93763fa7ada3f25fd299cdf54154011454a09c6,78dcd870a36869d98bba3e069be5ef96b7afe47143d5ceee126683bfa158bdb3,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,81.465,81.5,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,scicode,SciCode,coding,b4896f26497a83368338de672588cfb7bfa683553c9f564b7f4e52b59f1fe2fd,d9d3d6891dd5f5487d32f261c0deacd3dc2b9f7dd253d170776802c6ff74c733,644e6a16e1503037efa3824b666795244b991da92c77edea91a9f52f77fee232,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,52.0833333333333,52.1,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,ale-bench,ALE-Bench,coding,daa9971d878535bbae28bc7a1f5bc08f9e6e6e85610ee86dbf6584df4912096d,109e3f5a118e1f980cfbcacba3ee520183ea1b8e4bc3ee816acd46f6a3815f6f,5d3e6c572090ddf18d17752ee4d05457b8a476757b729a0a74b381f830b00dcc,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1188.58,1188.58,exact,canonical-source,profile_cd463b62537d29712133caab,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,vibe-code-bench,Vibe Code Bench,coding-agent,92d9636da8a6207c082a58657aba6f9f89d3eaa594c63f958fd0b4b4538ff68c,981e61728794cba1a13baa812f14811235d367b449329321fc3a703d49176462,304e82c65fab2b71f3b2e514b7ee483ad4399619ce2c6ea05928a5880317f8b0,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,47.969,48.0,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,webdev-arena,Text Arena · Coding,coding,67da42e438e54bb49450d2b8ccf775248639b61186b3daf4df91dadb5892bc01,1c0a731d64f3c34e2ec8efd1a6f0e89d55d4599635de6b3c0e4b95a0a5446d62,506c358899f145260726698d1682f26cf00080cc3d585b85a7e92136103dfac3,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1397.299194720604,1397,exact,canonical-source,profile_cd463b62537d29712133caab,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,programbench,ProgramBench,coding-agent,7f59c34b09f885b1a4f834f14d771a6e38b9205ce18982937604e3da0867b883,2c72c623a77a8ff708324314a44785f83e210d29dec4872bf855f776a0f11db6,cb64a34cb8c19214a8e30a3566309ab0e9b47df4eea36a9d77a3ef0800074d4c,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,da4b9de6ffc1104ae1de450ad48b6c7d1bf7d01f809037e6a7c61146ef37a871,642c100b11df2c526aeaf85f940c2370baaa86ab8aae8c32222824fa42907550,7f427938aecc10eab238c7bd96d09c3efa44115f71632c55f8a9f164f10379b2,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,54.682,54.7,exact,canonical-source,profile_cd463b62537d29712133caab,high,high,true,fixed_effort,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,8e57656f351ad8ef919b300b628ec3f7d850c257ba19d46d59f2906644355ff2,fcf1457fb396aa4f8a0260d3e32d91214b57b55c70cc3921db34443cf7e50316,c06fdebe841d5add776f6a399db64ea85c5ccd79a0be8fe889c197801fa1c124,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,2.525,2.5,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,skillsbench,SkillsBench v1.1,tool-use,5e27ceca0d7c694bf196c1c987c4d9833e418600e361a562e5b5f95c55bff6fe,0befa8ccd464ba5d0240bc986890964300ffdddbb51089f3b7886c9c61a9b0cd,c2ed2025e534826e5e627446f4493a128269710491b35eb2122abbaf19d56304,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,41.4,41.4,exact,canonical-source,profile_0ad87315d87f37ec5bf6e958,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,bcc57014069af24c8daaccb9dcd1029156cd341ba90645417294d44ccef337eb,216560cc378a7d801c764704514bcd71c87674a03acd2fbe9b2863379306c51a,eacfae557cc981fdbbb8a10002567bfe3e737e5aebb5da1e1d848ff3c154a0c9,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1021.79,1022,exact,canonical-source,profile_c6f0f8947c7f8dc16f5517b6,xhigh,xhigh,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-5-4-mini,GPT 5.4 Mini,OpenAI,closed,64,weirdml,WeirdML,coding,6c8c8288cf96343444ca8d9af0225d66e120586a25400e1a34955721b2d77a81,a99d9e66f8d3b7564b4274b9786c0a9ec5e239369a59a675a7780497586155fb,08670d47312fe006c5a06588bea2dbeccbfd7b97875e82e7944fc1ed29ca942a,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,60.3,60.3,exact,canonical-source,profile_cd463b62537d29712133caab,high,high,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,epoch-eci,Epoch Capabilities Index,capability-index,8dcfe9bbc12205e797fbc6ae9c9e3b261974e998711c65490bfa812306142264,8171608e91bade0728fc273b33a6928d1e09b60243753514a2ad4a494844c2dd,4aae907a1d796e85523324700c3de958e63e1cbdf9ab302865ac9cdc1abfb748,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,150.2,150.20,exact,canonical-source,profile_f5e834ff6a9d4d675c2c021a,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,arena-text,Arena Text,preference,ccbde14ead0933e2eb7a0fc5cca72a818e33573a4811ec38bce24a5646c37423,434c3719dae7f10e1444d03d98a941fa1dffc66cd49144782275408d9c8cf9ec,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1473.6904735629234,1474,exact,canonical-source,profile_130345efddb139a5f20d658e,high,high,true,token_budget,32000,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,livebench,LiveBench,general,5e2cee8e322e531eb7c7fb4d91a80c78c7195111a246f6f5858d0bd3efb959a4,1318b23dcadb7660f8e584baef5824865bac2341937e66cea62db137bc669487,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,72.58216666666667,72.58,exact,canonical-source,profile_139bf317a1ccf46d0ad71c1b,high,high,true,token_budget,64000,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,5925eabc44f4dc32c96dae1ab2abdea4d18c065049e0ed27cbff3fab05f865b5,73cccaadfb01660d28da0fed352f496c2689f0d9191c57e998d51db4138b9b8d,f8d489c12f499bbf0b15f0bac8891a07254268ab8d7e56a22d7e00cd23f2cff2,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,14.16,14.16,exact,canonical-source,profile_7c52e44fa1806e202f8c3942,reported,unknown,,unknown,,false,2025-11-26T20:25:03Z,stale,339,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,simpleqa-verified,SimpleQA Verified,factuality,ec8fadd030ac501d2a8b9e139d4e132d5283b1a16d9eed75dccb8ad776bc86b2,d569e03b8c346068e029a97b3abf6c71ef2399e46457cb306b49760c3d1ad5de,644c85525faccfd7fc3a26dacec1284e5908cc1abd3c0b92a30923675045f344,release_ccacd6f679c9142b534364b4,task-v7,verified,accuracy,percent,true,39.004149377593365,39.0,other-supplemental,supplemental-source,profile_a72b975b52d107a214a49b26,reported,unknown,,unknown,,false,2025-11-25T00:00:00Z,stale,341,1,A,1,A,0,A+,official-benchmark-platform,0,official-benchmark-platform,simpleqa-verified,SimpleQA Verified,https://www.kaggle.com/api/v1/benchmarks/deepmind/simpleqa-verified/leaderboard
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,gpqa-diamond,GPQA Diamond,science,5c283b1863945eeb1ced6280a618610bb3cfac94f2e65aaea86b9e29e91b5a4f,f045497f70a0d1c9c4f6fe0af0badc181d52f906fa30fd1172efdc87c3714d46,5d265cc0e54a9e3d776b2e0a2b8f33312db57cabdf2254c07e8691f3f3f37089,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,85.859,85.86,exact,canonical-source,profile_38e056a3f25f42f22b205781,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,mmlu-pro,MMLU-Pro,knowledge,e143addec52de82082d13b7287180c02abecbe5d857ab3066683b956a3b59d61,1bba8deeab7e3570b86daf5c87b584592de999eb1f9eb86e8624c24f2d23a3b2,d7e438fe3ddc56d7fbaa49bea102c0019cf3a472fc20942c8309f85c7c9acc91,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,87.26,87.3,exact,canonical-source,profile_38e056a3f25f42f22b205781,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,ef30b32555e0605330ebd0cea850634402408f610fd487e0d2445be9849ecb3e,d156fc4f0a15ab6edce11bae6ae0251444cb284c391df1c0c09292c5cc953b2d,2fa30ce7c948cb2db42398112d6401b364afa08c3e7b7a939b5a2d4e2f252d45,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,48.05555555555556,48.1,exact,canonical-source,profile_8163073ac76db4da90db1b36,reported,unknown,,unknown,,false,2025-11-24T20:47:49.364Z,stale,341,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,frontiermath,FrontierMath T1–3,mathematics,b4fbffc6fc6329da4aa514fb363753072e6c4ae1e49bcb78f3b19c269b1bbb30,48de2b49b95bd69b8a26fd9c2f13a24287dc9986a1eef79d15c71d7c62b141d5,25b36e77ff057a5430c68caa40fa10bc60073c75fd6fc13c93f34b7ddb269e8b,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,34.385964912280706,34.4,exact,canonical-source,profile_597cba506275a64a67c0cb7d,thinking-32k,unknown,,token_budget,32000,false,2026-06-11T01:55:25.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,arc-agi-2,ARC-AGI-2,abstraction,9991a1a186b1119daf0a6821e4f59d8bc1da79d1e0781d94ad58e323616f5108,5d0cdc4137e5eb3db8db5b9efe8c60655d11f87ee865b18c62bb07a1a5d1af76,4b3084c25aa997f35d6830e86f41d21548d7307ed80b7e834cdb86431ea7ed8b,release_c66f5b82bd3045b91ad807bc,ARC-AGI-2 current leaderboard,evaluation-pass-at-2,pass_at_2_accuracy,percent,true,7.779999999999999,7.8,exact,canonical-source,profile_926bdc433ee7f4064ce2e951,adaptive,adaptive,true,adaptive,,false,2026-08-28T22:43:25Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-2,Epoch AI mirror · ARC-AGI-2,https://epoch.ai/benchmarks/arc-agi-2
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,frontiermath-tier-4,FrontierMath T4,mathematics,035c206c331bcf243beea8d28a7aa2230142d7cabffc60387c97c23c30aa1f6d,5d9e57a0155114ee7b3a7bd35f6ba68d7e28bfde7c05f5d9751ccec3d1096c72,52c1ea0466e69627d2998ab353768b351763ae03145f03084de8f4c7c1fc4a31,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,4.878048780487805,4.9,exact,canonical-source,profile_beecacca1fd5f3ea74568a4f,thinking-32k,unknown,,token_budget,32000,false,2026-06-11T01:55:25.000Z,fresh,143,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,mmmu-pro,MMMU-Pro,multimodal,f2d75c69d326d564167f05fa2d77216641c254ca28ebfd91c618533d4aa7bf3b,ecac4483239faa9e514ce664152be84a5cef899a7dadc7e68bf52cfb7b8c84fa,1c15fb011e384b8a984f6a69c456c68cb1e57456cec7fc37c03ede17f209b136,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,82.948,82.9,exact,canonical-source,profile_38e056a3f25f42f22b205781,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,simplebench,SimpleBench,reasoning,c38c5b3ad541eda02385e92bd62a17ace6f6812ff893b1e6ef1e99ae56844ffc,7c82d671b3b48c6df64c46fa366eaa988c12b110787f968c5e738c3139d65e39,bd4a0e71c7f9c39ccbd5b3fab0cdc3be9fce0b80743bd8d092a4ce230a25f1ec,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,62.0,62.0,exact,canonical-source,profile_56fcc66f50d70a253236a8bf,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,chess-puzzles,Chess Puzzles,games,34c83daf144bdf4cc40197209321ca63c681cfd5a0431f3841dd8bee63619e9b,ed22f0ebf09987c064a3fe872e60ff5d2e7ed8454913abdd5905d78d7669c921,4afd2c28aad2fa3a03ac5912780e4205a03bf886063f351bfd2a27a4321a1f9c,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,5.0,5.0,exact,canonical-source,profile_6eb479ddca0f4373f2bde9be,reported,unknown,,unknown,,false,2026-10-13T13:42:24.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,arc-agi-1,ARC-AGI-1,abstraction,6801a646d6cea052f87352840d9f0e63e44894fb841319d270abbbb907c9785a,637e791d07c9939199fa668feb7e5b39eaa89f389fd5317b5acc86702c3ff984,bd170eff40b20afe8d63cbc3182879277ef4f6968b7f55e0284fbc36322a02dc,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,40.0,40.0,exact,canonical-source,profile_926bdc433ee7f4064ce2e951,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,swe-bench-verified,SWE-bench Verified,coding-agent,2bdc81db69bf376c7ddf15e96505709518dcfad0a6a6adfa0c923b406fb35b7b,9af256aea80ea0816cfc8cdc697148f716a7e4cefea34951c9ecba3d832a4175,266500a34989fc6bc3ada17acf30da1e972378b6bad797583daf3c0cc52735fc,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,76.4,76.4,exact,canonical-source,profile_38e056a3f25f42f22b205781,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,swe-rebench-v2,SWE-rebench v2,coding-agent,b0db892ecc6b19ca4db9caae151c345c3bf55c06a05cb6810cd972b8f021396c,14da9d4fd1eb38bd417024f7698503edefda5bf54acb32187b53beada1dd2408,55c67f29f9caeae2fb1a7af296decbc5ef7cf3d6983fdb72f33fcc65c4733416,release_1c18f4398e32b2e0fe7f7d86,rolling-v2,2025-12-01/2026-02-01,resolved_rate,percent,true,53.541666666666664,53.54,other-supplemental,supplemental-prior-window,profile_6d7d2d1817324164a6889f68,reported,unknown,,unknown,,false,2026-02-01T00:00:00Z,stale,273,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,gso,GSO,coding-agent,c62b12a8f515d8c85b08d0e31051d883fc5d6e8aa0e42524e15b629d1a864e67,b7067fb3bb3d809f77604b24b8e3a2bf0b8eb6cfa587f268b78e88d5b3b5ede4,72b1185f7df85085f15496c403080aebfabe615e1f401dd190b8f8a25bfe10a3,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,26.5,26.5,exact,canonical-source,profile_4c7c0b6e9b003d3102b8548e,reported,unknown,,unknown,,false,2026-09-02T09:27:04Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,livecodebench,LiveCodeBench,coding,56cf58fe02ed0ff933be71f6cf9b8818d059817368c4f998ffaae52902a557a7,3dc9933d1db16259ab8126292b7af1b5dc4f7926c07d94201c4f329da9e94928,cecc93e7118cb670afa1c1a74f65be3f0344051f547a97e5f392ee113daa0fa9,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,83.67,83.7,exact,canonical-source,profile_38e056a3f25f42f22b205781,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,vibe-code-bench,Vibe Code Bench,coding-agent,163e8d953ce876eba610c5e399db370b0631faebd918639e52f3770fdc6b59e4,8f938fcefcca26c9b6283ff64cf2257d9d326589ab62e5d8c3e20015f0f8baf1,13b83b290f4812953b87168bb38d8d29e562a97c32eb43cef14902469723aeef,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,20.63,20.6,exact,canonical-source,profile_38e056a3f25f42f22b205781,high,high,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,webdev-arena,Text Arena · Coding,coding,20db45869a4d6c0096d3cc58eb931dd65e0b3a3924fa29ca0450b9900cfc9271,7224f66f1d3fd2c5878d03ff6d6fd23b3b0d9759ed8bd6aa816b8f8c631a344b,b13d87ba71e682faa1a11073183a5b6282d1460aae5fa73c48679d9467e3069e,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1493.6607664512705,1494,exact,canonical-source,profile_130345efddb139a5f20d658e,high,high,true,token_budget,32000,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,weirdml-v3,WeirdML v3,coding-agent,35403a5961b65093af7f8d353bcd52dd8f4cb8d1ec81a3b6e3c42cec8a730c14,5460220703692c88b32a10fcc9991bb6eb71c91f66776a779e07c0e02fc1e63d,1a3a2749f388bd22e3e4ed0cd91ccefd02106aa3944e4cb9c223709565c62183,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,6.5148765151515144,6.5,exact,canonical-source,profile_38e056a3f25f42f22b205781,high,high,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,bfcl-v4,Berkeley Function Calling Leaderboard V4,tool-use,8d22944d7d361759f06cf27568ea3c19c9cebab1ef34ee5584a9f1ecb62ddf08,b16b927b0fc6a8c5c57ac2a62bb16f71d47b0ce143c2e4603aa9c1a7cfa7859a,6a9c99bccbca11748bdf70a2bdd6eb864726c94bf4a029f84ab95898fe96d192,release_7d2560723dba9d549cd10b25,v4-ede5081a24bc,overall,overall_accuracy,percent,true,77.47,77.47,exact,canonical-source,profile_add9f33c77552c619900f72b,reported,unknown,,unknown,,false,2026-08-28T22:03:18Z,fresh,64,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,bfcl-v4,BFCL V4,https://gorilla.cs.berkeley.edu/leaderboard.html
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,vending-bench-2,Vending-Bench 2,multi-turn-agent,c25593b775fc227f7ddf3e5f99bcd661d01ec118b1d3577cae3c1fd10ecab3ef,ae472a7b1b7fdccba65be59d83e8913c34a91f9e7af41603c0c7d3c99f8f1da3,d0d6155566171de54f7bb1d6fc1fbe71b75e8ef263810485607e592935427ef8,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,4967.056666666668,4967.06,exact,canonical-source,profile_acc9610846795e84cd094615,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,gdpval,GDPval,professional-work,e82ea31213a709ffe2cb81e68deca2fa1f303845cfe2797d37e42aa0f93872e0,2abf0ade10240f8830d9f6bba725ca3500dfad78cbda608ea6ea6d11d5dd13d6,378fc1a8e117ab4f314c4ef7da05eccc63a48ee1c23b7e95a30acb27f82c58b1,release_730527fa346470703bd19e21,GDPval official leaderboard,expert-comparison-win-rate,win_rate,percent,true,45.5,45.5,exact,canonical-source,profile_b5b92874a5c47d9cc6e62f91,reported,unknown,,unknown,,false,2026-09-02T09:27:05Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-gdpval,Epoch AI mirror · GDPval,https://epoch.ai/benchmarks/gdpval
anthropic/claude-opus-4-5,Claude Opus 4.5,Anthropic,closed,65,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,60354dfec2ad4573e95efffb3135c8c8cf0e63a452855134cec3db53022ca2ff,9a3d4866699a04b9f9051d9796a8c11ef3e2c47a4d17dbbcb995cc7de09bff7b,67f26e9643b98a8dfb5466fbbc338423afd93b9175dbd49cdf6dc026e61077e4,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,24.74,24.7,exact,canonical-source,profile_38e056a3f25f42f22b205781,high,high,true,fixed_effort,,false,2026-05-05,fresh,180,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,arena-text,Arena Text,preference,eac64dfd7af3fef51ef4d399d066488072d9b392b9c3952bbb966aa0b5442165,0e6abef28af512d7d30f549bc9502ff7bd802875a4f276f19ea7e9cc1fa41061,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1466.599824571988,1467,exact,canonical-source,profile_cf88c4c74c23c65e8505ad74,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,decc6b5e09456be99db61bfc3196dac55313bed2c73ac3031a3cb7a74a069bb0,8bf66fe5ead5c0c61ca5a7b84a0901923b58ddc74248303b3751dc3b5f234d95,9bd3fcd30506d02bd2fd6283175199d72d7010226d22784e554ab0bbbf43b033,release_5f3023448c2ae4632fe1a43b,reported-runs,public-multimodal-mixed-run-configs,accuracy,percent,true,33.8,33.8,other-supplemental,supplemental-source,profile_dd5b317ecd1d74199c314ed7,reported,unknown,,unknown,,false,2026-04-22,stale,193,2,B,2,B,1,A,evals-report-verified,0,evals-report-verified,evals-report-hle,evals.report · Humanity's Last Exam,https://evals.report/benchmarks/humanitys-last-exam?tab=scores
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,ac1f216b64fab979a2878990c0077d5f25cb086bdc5cbeefa2037753f77fcd30,c0a8657a372f1d53876e15d69027514d3ac33939aa9f56bc91d774a7e7f09ee3,cd0461dadd329d0be451543ce165ca81ac524d86875590417582db8b5908c511,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,35.681186,35.7,exact,canonical-source,profile_8fe48d99b25cc95375221a5f,reported,unknown,,unknown,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,gpqa-diamond,GPQA Diamond,science,70bee2b0999ec5cab4b4092d798b0fbb1d080a75a25718436f52677fe4c35aff,05e7a854fed643780920a931cbecd8feef11821de204f3ad784ba317a8d22f32,3809925a5979fd5445613e7b58fa86405fb820e071fa34c0a1624164881ca6be,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,82.576,82.58,exact,canonical-source,profile_e8c081afa28a4a38e26e8d28,reported,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,mmlu-pro,MMLU-Pro,knowledge,e79bde44ef42d1f0fa344b24e2b352bbb8e76a5aff27cae3e84c572864ac1cc4,b367f207f03d36860f039bb02f4ff2fc697638642d507431f91487afefff424d,8e9a33c426fa64de82f744a067d698c830971c2fbe2c88b5aed9149fcb068a0c,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,84.595,84.6,exact,canonical-source,profile_e23668cebc6abd250154d2c3,reported,unknown,,unknown,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,critpt,CritPt,science,aba31ead9292d7e54da2ede6b513fd9913b2b27019033925f7feda3f13258b25,cbc911b2c0a893534395eba203ed18d6876599c1f7dfd997109aa7f8efd2acb1,0c3ea6fcbb859d82df8441105ff8a6e25c282fc65d419aae0af8d4f2a9d33f39,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,4.0,4.0,exact,canonical-source,profile_8fe48d99b25cc95375221a5f,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,swe-bench-verified,SWE-bench Verified,coding-agent,3b228685cd607ba15edf9348b44da33005a7ad6274d914f9668e3f758f116190,49bfe323a7a9b888f89260f511f0f9befa3e0bf6f40d27f49aabdefecb734796,864652b56c2022f2d0bb0f0d1ec0399044a9c8fec8f1fbfd2a76e5a38479aef5,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,74.0,74.0,exact,canonical-source,profile_0785c90f25f84cb66ec81391,reported,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,swe-rebench-v2,SWE-rebench v2,coding-agent,9c37c659f201b32987eff894cfee9fb41b5a9c32ae19a45c2474faab2dffa471,cd7e714afccb62b9c0e75bca9a25f10c60e2f378e0f231976a7859a55918de3d,49ad2dfb26d0e8051efe67dd32de507e6b83d1a5fb53f1b6ece7b758fbea33c1,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,46.48648648648649,46.49,exact,canonical-source,profile_1fb89a7ed9a95fce8391c709,reported,unknown,,unknown,,false,2026-07-01T00:00:00Z,fresh,123,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,livecodebench,LiveCodeBench,coding,07a8f0bf821bf964244eb79a50d5d81f378c390052afe79c05d29a0bfb3690b3,e7d887be191c638dc3efbabc493d631a44ea5b400e57381b871e4a00f15ef9c6,0711e0cbafc33a43a7c9043f2ab727348c193386d1387c59f9197403556fd7cc,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,81.354,81.4,exact,canonical-source,profile_58d95fa53ab7d749ad91dc8b,reported,unknown,,unknown,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,scicode,SciCode,coding,3c54ecf29e6a30e3cd12b552160bbdf2d05ce6a4ca4f9beed8124ee58505baf9,847fbd4846ca3b3e13a3017b757d337ec53024e4604aef10ee2a4f4ec47f7341,e77e371acf4336f05f77ff67729676f9e7924a5ae6d4584134e07e7d70528dec,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,50.2314814814815,50.2,exact,canonical-source,profile_acb46f5b3108016822f0c8e0,reported,unknown,,unknown,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,ale-bench,ALE-Bench,coding,965f507ab05c1b0f28e79cc03ccfab08248fe5709ecb9db40385852df1d38895,0c6bc7f237aa84b101b6ba38b86731ac1fb9f5947093175f67dddf0d8ab28366,332e5148a14ee8714a7566d28a9eb053d3e7fa4e32accba9de0c5dac596f24c9,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,899.8,899.80,exact,canonical-source,profile_f27c99ca4ea89e5927d9d756,reported,unknown,,unknown,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,vibe-code-bench,Vibe Code Bench,coding-agent,f1903ec7ef5f7b5c6ebbcf9681a683aca8c000ca5afa2ef0e8469f956390ea76,0b171a0387fcb7b32020d50c57a607cc7cfd01af365f4533e86bf449a2301e83,e118ec95a6d19aa65c4107d4979701f48356cc6508659b5054017d4dcc9f3b1e,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,34.113,34.1,exact,canonical-source,profile_33eda8b12928b09184788407,reported,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,webdev-arena,Text Arena · Coding,coding,fd35d6d7e52b44c3f770e1e24bf9035352154d28bbda00ce56028a652c6f828c,b993fc30477771ba347d5eeb0002a35a6d851b6d6a452de761e9fb02c0078171,46b722f1547f93e3f68f5238cd32f974258d8c8948ac6d4c68d410fd978bb37a,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1480.4355329379657,1480,exact,canonical-source,profile_acf227e528e7a25cffba9cf6,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,bc3064a201d471fd8b8909cf4a25260efc6a68295e8ec16841e1fc02cbc68d32,0bad67720ae821797a6148e0a57a51eaa287c2cedfc91a3562bb21207e72c197,80022cccfcf77158be6f9c315263156b66d144105405995ab53279ca4f158761,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,57.303,57.3,exact,canonical-source,profile_6690d5e0f6e56e1612dfa415,reported,unknown,,unknown,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,15819126fdeb1109683f242097ec2c563a84cbd6c9424f7da71d79934eb632c9,d2b5bbb7da1c4ad2df44753b562d9fc244fdcff99344bdd6d5895eb65bf3b987,ca84380a4024d52fc7d440451981caabab206a341c119e17c9f60894bedba933,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,0.505,0.5,exact,canonical-source,profile_8040fea5627284b131560e23,reported,unknown,,unknown,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,9b5d2b8397c525d2e3f5fb0260b620f6283b1bf60b8a3a43b7da6978d5e621e4,ed630d4f7504cfb817a5183556f0a89e03a2679b5940beedae0d091a7b0fff20,66241d651244d29c262a1ee61a7e597df71b9eb9e7f7dae6ccb808ae74e61581,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,9.896907,9.9,exact,canonical-source,profile_8fe48d99b25cc95375221a5f,reported,unknown,,unknown,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,arena-agent,LMArena Agent Arena,multi-turn-agent,54784f0db036b973cf4cc1b442923482ebb48c84b80e23a0be65e8a9058b559a,1b0f5718367bb9d8c8589bbef3a56ac3276129c7f64bc7fff4cd83ad838a0b4a,f70df5b9f0aff58ead73e129c7080cc3691c0764a7d15fbebd11a30f682481b6,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.08932691327565813,'-0.0893,exact,source-specific,profile_d8ac0691b0cd4bec10937bd4,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
xiaomi/mimo-v2-5-pro,MiMo V2.5 Pro,Xiaomi,open,66,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,977f8ded466eafa646eb7d24a3c84aab7509d29f8ff0bdc9b0ecbfa596aa24a6,e65401b5ff62d089583bd0b61dc453118ffe489c72e0db522e92c53257fef441,609edaa4443866ff024ee88d9c3acb2f02e7f9e00c012b4c9188fdf4d4ff1f3a,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1128.63,1129,exact,canonical-source,profile_53135ec52b36e5ccd955ed0c,adaptive,adaptive,true,adaptive,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,5f8722d88ab0f38e798786c92b4d830745db89f24a43fde25ef4c0b7cee147ac,aedd9b98ab97cefd41ace0b279a2b0e34d1f3861f1f40990b6cd67c7cea2886d,5b7e3eeff2eef9ab59ebc6432b2bded7c577fe9bd629cf10ded96f6b139cbfc3,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,28.528309,28.5,exact,canonical-source,profile_6163e1e4e2e25c4ca85b2f88,reported,unknown,,unknown,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,arena-text,Arena Text,preference,1bf8412fa0ea7e29b5bc3a7a0a44f7df84be7d1c44ecdf78cdd0edc624b6a499,f551b07744bac5f63daaad9885df9cd2b0e33bfae8162bd0db09d719a92f5542,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1448.3740011611799,1448,exact,canonical-source,profile_9f6dcc62a58559d366b17164,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,a3366db8bcfea21183b1c52235fb237887b92bec998ed195e5b88fc6ffce74a8,b8c3cde3423d1a89b7d8dcfcc3294b6f90a8d0d1aa4eae4c4ae2f8dfa649ee64,dcf90562187180c47145fec344807665f6ec83362938f3547fdef2b3c17b40f0,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,46.478221,46.5,exact,canonical-source,profile_6163e1e4e2e25c4ca85b2f88,reported,unknown,,unknown,,false,2026-09-30T18:53:23Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,mmmu-pro,MMMU-Pro,multimodal,81bc742ac149236b1d99143b7626bbf7028df4cff08387bed8e49e23d1015ec6,4350372ac3dc92ff71163c4c13cc27e7b793f3637973ea62436d7e8030784754,6b757dbe83d1134f358c6084fb236e80413b45947fa8968d75244506ad7ac8e1,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,76.358382,76.4,other-supplemental,supplemental-source,profile_6163e1e4e2e25c4ca85b2f88,reported,unknown,,unknown,,false,2026-09-30T18:53:23Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,critpt,CritPt,science,7f55ac64f00a479542ed801832541c15e6d9d213d929e5a3aac424d7a0a94e50,c2012c1e5f904113d03427dd0150fc24d7a50e0770e34d95f2640dc92464b9d3,53d9ca1fce38485bc5d138d525725f6f7fdd90f6c5436209d856d125e2da461c,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,20.857143,20.9,exact,canonical-source,profile_6163e1e4e2e25c4ca85b2f88,reported,unknown,,unknown,,false,2026-09-30T18:53:23Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,ioi,IOI,coding,0bd06382806a98c8ae1d3404291a714bb0da39a5da6d1998b08a51054fc2947c,8f2b52c67ddd105f75291254359dda62ed23edcc5d039ae8cbb4beb28ff28990,369a3d45b774fe21c4a88153ddc53c7161c10fbd09f232e2acfcea71701bb31d,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,38.056,38.1,exact,canonical-source,profile_80831aa0cb0d312c1beaf624,reported,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,scicode,SciCode,coding,a2827c2706ac74c3c7527a1364f8f03f4e8ed03aa9558d3e0d3a6cac1de6fce2,faabb805206e8b25a16d375a95665bf2f96057500b134040e442c86d32cde29e,962f818b99d0ed7640b7d946bb09198eb356bcce789b2edf6bdc6f821c7a3019,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,58.912037037037,58.9,exact,canonical-source,profile_9844486f407e9d5626c38e1b,reported,unknown,,unknown,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,vibe-code-bench,Vibe Code Bench,coding-agent,3adeb561eab52844360ad2fae57a2cb3f43a3e25496b68a7f1256d620d57afa6,1bdc690dc97e33172f043f720d68749a69c631f8e0775510ae5affc98c1ac01f,c9312e0fddbb682fbdc5a2c4fd70cd9ea96ccffd05a50ed21acabbfbc7ef5fc3,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,70.399,70.4,exact,canonical-source,profile_404128e4ce18921deeb03b6a,reported,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,webdev-arena,Text Arena · Coding,coding,1be3fa9ad7c7df12baa3fe24712c023659eb703f5025ac5b2003f344593cc479,94374aa42b4385e8e0247b0d81746281d7bb77880e929e86536d1c378a063833,2abf9de703f38c4eccfa33b5d1cd6f4d331467691d3245db69a479e6ca4f6af1,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1571.03868087871,1571,exact,canonical-source,profile_eb70fd26c572537944129f9f,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,5259e80521341a171279df007ecdf22928992eb307c6dc8e961fff52fa451d08,b84b40e198607408a333845b1a45f3ec46d5e0d5d5d05b450d865bed71ffd4f2,6d46feb70a5389e6949c036a71dfbea659fbcf1def3e8db0ec0415f044dd9dcc,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,31.818,31.8,exact,canonical-source,profile_81929620d2e48ed2e1123532,reported,unknown,,unknown,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,sre-bench,SRE Bench,terminal-agent,06c623f574c419d36c31c894caa3cda9223528d79649cd5c0f6af2daaaee75a1,b62a3c51f5f1ed7590609511522ed84fdd14f1199faa4322808e6337225910f0,a0e657cf73572616f8deda4e80c9873148b55876068435e4c6e0596077de6ce2,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,0.382,0.4,exact,canonical-source,profile_cca308f9e01203fcaa3c0965,reported,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,arena-agent,LMArena Agent Arena,multi-turn-agent,59c0c49cb20538dacf01ceb6a875e2fd753bb75dbe47b0f9adecfe64271e5c7b,ea20f19f811fd357750cf487d73241686a677e452918a9967f9d7f674483560b,017f316bed400a9a52ed55035fbe8bcbfe9a3ece72f6a33b2f6e99c4e2595b18,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.026798104156421143,'-0.0268,exact,source-specific,profile_73b151451dd1e7342d7ff672,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,c21c7a4483d870bda24b45dec7dcb1cc24469a91219eb15713b24f7244dc5575,ebe293e6c6a4af63f3e96ef7cf9874916ebbe368d4ee0e9f3af548c9e2f86b2c,1dfa9cf9491fbc336932063c2a6b54081c62a4537427d22ba3c521094f8286ae,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1593.99,1594,exact,canonical-source,profile_6163e1e4e2e25c4ca85b2f88,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
stepfun/step-5-preview,Step 5 Preview,StepFun,closed,67,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,218250f99c79754fdb7e8d0fbddb20a6108c8482e2668829add0fea10e0a4a7c,3ce8bc3969e35cc52967474bcf9e85506907699907bf234fcdeacf535a5489f5,39d88bad143f94079f2984645a9ab692b81d3d240b2832fe90703029f137ebad,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,2.857,2.9,exact,canonical-source,profile_0420341c0f15397d2dec1ffc,reported,unknown,,unknown,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
xai/grok-4-3,Grok 4.3,xAI,closed,68,epoch-eci,Epoch Capabilities Index,capability-index,497ec008bd1c017740c4ec8c0328efa65f9764fdcc6c89aa349b57d9065ba62a,d2d0478dfb32e6c2242411b628422642fb84e301b5c9940f7c148171fa4915ab,6d1afb6719ffaceafe3641009407d50b145d3c5e3c7ad7fd08a6f1003ca70b20,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,148.95,148.95,exact,canonical-source,profile_59e6bbd0b1c44e7d3a4db8e3,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
xai/grok-4-3,Grok 4.3,xAI,closed,68,arena-text,Arena Text,preference,ea18b1e035ddfc0960a392eb8fe20d5ec009e43e9b9aeffd99b08975d5465cd7,6bbc2652e37e560055fa1e201c7f976b2e4718f41276c9a227d639adc9d1d8a7,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1441.6307836382434,1442,exact,canonical-source,profile_3c0b61d8f0225024eaa77540,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
xai/grok-4-3,Grok 4.3,xAI,closed,68,livebench,LiveBench,general,6aecd69e8240e7acd9121fc9ae9116cbc5c02b7f3603beadced7d6883bba54e9,5b1bf78e46e617917bb9e4ff6152e7a7103c28a7b526832db35ac47d668757b3,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,62.24735714285714,62.25,exact,canonical-source,profile_696ee2e03f20836e5b0f3f47,reported,unknown,,unknown,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
xai/grok-4-3,Grok 4.3,xAI,closed,68,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,c630b7051d48f2911cfae02c145a3332b085e2446bd5df0032ebc6bb79681856,360fc8c58b83e4b4a9c6f42d710d9098d81653226153dfce86fae730a4205acf,d0d156c8a19ab9488a31e274e6e36ac05b2a97e639a4aeb7a6342e19708aaf5d,release_5f3023448c2ae4632fe1a43b,reported-runs,public-multimodal-mixed-run-configs,accuracy,percent,true,33.12,33.12,other-supplemental,supplemental-source,profile_d2089c665b18a125cc91267e,reported,unknown,,unknown,,false,2026-04-17,stale,198,2,B,2,B,1,A,evals-report-official,0,evals-report-official,evals-report-hle,evals.report · Humanity's Last Exam,https://evals.report/benchmarks/humanitys-last-exam?tab=scores
xai/grok-4-3,Grok 4.3,xAI,closed,68,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,7fe47ceff9ea227275f26c61fc0941a1779ad21f767ac30d13e15b5cf1472021,f61bb65e81d1b4406a2ac544a8d0ba50ac3639d4ef02152c3ca82b45d59ae7de,eefdd6e9cbbf4533f92c58fad483574dcaf1f14ad34508ad557dac64f444dbdb,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,29.981464,30.0,exact,canonical-source,profile_1c8c583aa239a9f12d2052a4,medium,medium,true,fixed_effort,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xai/grok-4-3,Grok 4.3,xAI,closed,68,simpleqa-verified,SimpleQA Verified,factuality,5cb539efdd102e2d1a1934c0a6d96b9b8dfa0dffed0102faa56b1ba1d6f98bf3,05716c1f1dfe5591d015f25270db15ab6d20616dd62ba04387e3fc8b26a1f4b4,7d138963618e04aa90455cf99086e6be055208c460d08e4d47a196a53d8b3ae3,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,33.2,33.2,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-08-27T19:29:33.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
xai/grok-4-3,Grok 4.3,xAI,closed,68,gpqa-diamond,GPQA Diamond,science,8408f9e21619e5e168d7e1cec5dc44f5fcb6d04076da7ac972e1c781444551cd,9c0e897eada1d59bdfb15609425bcc5c0338ec87b62dfbeaf95ebf31868a5969,4b68b93d0d9f71b5644d6090c793acd555c4ab5b3bd06fdb8242d45e3991d9e0,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,91.414,91.41,exact,canonical-source,profile_e6219133f7d970918eb71d3a,reported,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
xai/grok-4-3,Grok 4.3,xAI,closed,68,mmlu-pro,MMLU-Pro,knowledge,a685eed28666d1c36e8ee18c90d5b0dd9068574f8612f63cf34736d0976ee85d,18b4129fbb8fdc5101dd721f3ba0e3c42b5c4f859c44299bcfb9453510c0be0c,6b2a3bc18d23bd7aea07d1c6510f91c4558c5f54090d87c2da3523bd19ce9ca6,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,85.838,85.8,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
xai/grok-4-3,Grok 4.3,xAI,closed,68,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,02a623af9d34c7fdd4401ca6dd2423000bc41834dcbd506c8db07243ecaae21a,4bff4f49db8650fe0d8e6658c71b34feb8db9f66cd918a11ed8486d463cae4ff,249b023e199255073505beb6a67e06b60015ca533978a6656bff2f43483690e1,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,93.33333333333333,93.3,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-06-17T17:51:54.000Z,fresh,136,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
xai/grok-4-3,Grok 4.3,xAI,closed,68,frontiermath,FrontierMath T1–3,mathematics,bccc52dd66d20e74ce573005584e52e9b4f2c71504c7bdbf2a7f2b3216f82e07,dec84cc91bc62b346a72c1490916eb798274a080c6e5542785e98f4f3bb6c787,b957245226b13d90d28b7cbd6831a61d6ce276763e5442919988d1d8561d271b,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,42.80701754385965,42.8,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-06-17T17:10:48.000Z,fresh,136,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
xai/grok-4-3,Grok 4.3,xAI,closed,68,frontiermath-tier-4,FrontierMath T4,mathematics,6386cc19316007c40e522a741a7be9a85edde7862ab6abbcc46400b0b1392ccb,f64418a6589093afd1b3a30b1e57d148176e6aa96741ecba28fc7ee819a85107,07e2c15de85b808280d356662611df8d1b6699a3abeb5bc117715cef545aed89,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,14.634146341463413,14.6,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-06-17T17:10:48.000Z,fresh,136,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
xai/grok-4-3,Grok 4.3,xAI,closed,68,mmmu-pro,MMMU-Pro,multimodal,9bcf0579f7cd262f7a8edac4619900c94eb8761c5454fddd74d25f066cafe196,2b12536e50956654118c2ab83d5bb749c36df968c195f95fda720f30eb5bd2c2,fe465ec518e4224ab3918fed0b4971447469d1fd105a52da216f494e4c47d9c7,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,83.064,83.1,exact,canonical-source,profile_01923756939b60f21f059b03,reported,unknown,,unknown,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
xai/grok-4-3,Grok 4.3,xAI,closed,68,critpt,CritPt,science,0b9737e36142d00c7d3e9282bf2bf1304a814420e1acf339f2302d738ebd8352,a734266e2aef9b557287e7012478a2328247549ce6202a171f659d3cc78894f3,532daf285be09e39d56c072e92ae6b14218de708d839304b6ae40f1138330f0a,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,8.0,8.0,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
xai/grok-4-3,Grok 4.3,xAI,closed,68,chess-puzzles,Chess Puzzles,games,3e43081b46f624d7f4ab1d9c088a54fdb174b433a4d7ebbab37c81a334718a58,1bda84316796df226f117fe2dc0e6b3332f4b342e8be880ea80bd7c477f76dac,7498cdb388e0f524a643f5988aed656ab57e66bff68d4818eed1961ccb13823c,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,18.0,18.0,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-10-13T13:34:31.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
xai/grok-4-3,Grok 4.3,xAI,closed,68,swe-bench-verified,SWE-bench Verified,coding-agent,8cb0d7ca73ad5a104e355bf9f5df33a17daca84b1dabfb87179da2656830b707,fc3189a38d72abdc5e53ef1bd99a0196b57f60d3e59d721e079aa2ef6a2d0aab,05f2f7c4256d37844f23af9f6c45dadf2965a17e3370401f36db17a53ec5bda8,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,71.4,71.4,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
xai/grok-4-3,Grok 4.3,xAI,closed,68,livecodebench,LiveCodeBench,coding,50a5f174da125725b79745897a5f1cb44386cca0dd26dd5b5da951a863a91b21,b80813fe58bf3345a261564c9f1add521d0b9c517fb1bfd85c18b1b76535950e,e7347e25f2230e98cc2980f50512fd159badf6ca777985ea444e5fe9086f1838,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,84.494,84.5,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
xai/grok-4-3,Grok 4.3,xAI,closed,68,scicode,SciCode,coding,e781d330058a7ae5162b69a2706b936aa7b69a400b4ca7a926904d9369a009d5,82d2d1a6cb6d11034646d35c09db289b5667657ddbc10aa2f55eea60bb147deb,71a10a968bbc9415c02426c9ecb3c26f0e7357061b93d359bcf5302b77d39529,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,48.2638888888889,48.3,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
xai/grok-4-3,Grok 4.3,xAI,closed,68,ale-bench,ALE-Bench,coding,60fc1de72b26791163f0a4aa634f0fb6627838c8af0ff09e8141862e33660868,2e38f38946f4cc91b4bab2288d401fc3b5b779f8b862a51527f73d88528672cf,5c52cef2c13349fbe526e4444028f3f8d339f3aa4a2cef822aa8893adcb1d572,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,944.17,944.17,exact,canonical-source,profile_43f75a9097da6dcf80b67be2,reported,unknown,,unknown,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
xai/grok-4-3,Grok 4.3,xAI,closed,68,vibe-code-bench,Vibe Code Bench,coding-agent,02e3723848ac242e157111cdfc3336cd9a19dcc7eaaa9e7abc20b043e87aef5f,06a3f30fb2cbe6965793834df09d861131e570d7ee5500872634627d331b4f67,204cbc04af485f500377c94016fd006a739dbf90c5c43039b4003500b74a4589,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,19.403,19.4,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
xai/grok-4-3,Grok 4.3,xAI,closed,68,webdev-arena,Text Arena · Coding,coding,b7f24910585f93f73791a4fc8f3fa8c61b702f4ec94335f96be1cb4943290be2,a2672a7b3292e5650e8adc663d961953449954337e8b72eeccd068cee12a0e07,d59c63c4ff5944ac07f08a439182eaf5241f00ebddd8932f323a2b4232ac1f0c,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1357.5695879962777,1358,exact,canonical-source,profile_88f0aaf5c213ac8fa693d6b6,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
xai/grok-4-3,Grok 4.3,xAI,closed,68,programbench,ProgramBench,coding-agent,3a9f7e32dc018bfab33f24d594ef2787f7cbd2b735a607d2817f2108cf955b57,63ccbd375a9635092a9c75d93d748903ac2dde407ef309ef242aeee9863f42e2,89f16a602e79fefade7de87d90c75548a0f15b2c7744d0c45e83027c26bb197d,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
xai/grok-4-3,Grok 4.3,xAI,closed,68,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,6e06b64580915813b127ebb480166bbfa0ef3f77de609482500ff4a9e882d2ee,05e25410ce508f0c1b07046009a2d6f6f7fc2adc1145e176f1042837f4026204,d137423a0a3d6a26a943a469a87ef0a01a4b82cdf1ce9fd345d6428c46a4a778,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,41.948,41.9,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
xai/grok-4-3,Grok 4.3,xAI,closed,68,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,396dfac7fb6107ab6212cad4f32926c44c4d1ca2ea675276ed3177c4b4f763a6,27b6d9605ca3c250fede56c8ed7068d55f18ac8c0b8e14dc149b117c01941d07,5979f397674c22e4d08500584a4477e8b0036be8b47e06f193d5f083b6fe908e,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,8.041237,8.0,exact,canonical-source,profile_391f84c7bb85d165acbd4de4,off,off,false,none,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xai/grok-4-3,Grok 4.3,xAI,closed,68,skillsbench,SkillsBench v1.1,tool-use,4abd2f6114e6b010340b7aed1f880b04e2739459ad08551eda023a56454c5dd1,57f60ac425091578f669e9f453311ae1d584fd6086bd3432b344bab18c4ab785,c2ed2025e534826e5e627446f4493a128269710491b35eb2122abbaf19d56304,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,41.7,41.7,exact,canonical-source,profile_943719f03cd41fdd64848292,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
xai/grok-4-3,Grok 4.3,xAI,closed,68,vending-bench-2,Vending-Bench 2,multi-turn-agent,814e467bd261d8c2bc0f6ce6907a476aab1d5e8a9a0e81df3ad824b9e73a1fab,160041ab370d6a12c15c955b821d7a073e1737ab7831c5581d65e507faa70cf5,d34d1ddb40f3254c5f4a21bb94521538b32d1bcefbbe4a019858898e629aef75,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,35.26166666666665,35.26,exact,canonical-source,profile_10dc674a660bf5857afb18fc,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
xai/grok-4-3,Grok 4.3,xAI,closed,68,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,bffc57eff98f789f43c66db20d009c4d899d3181f9a0257effa8280e7589e565,860fde784ecba1fbbb6c8d84e94bcac910a25f90d5fbce772ea594fa657225bb,16bd08254293e743dd37f0e558c960f648550062c3ba6cc8a9b8fa980484713f,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,945.75,946,exact,canonical-source,profile_a8d922f83837348cb06cd2ce,high,high,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
xai/grok-4-3,Grok 4.3,xAI,closed,68,weirdml,WeirdML,coding,af07d1102510098898f80f75f379e8cd20bdeac356351647dcef45b360039281,924bf460811680cd957d0e42ae4060b4a39b3f74dc59a743aa83ccada6fba782,6f35a6ba42aae1245d3d9ffefa92a51ab123b07abc7377cca91a179998927e5e,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,49.89,49.9,exact,canonical-source,profile_7774caf4dfcc1695a2cae560,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,epoch-eci,Epoch Capabilities Index,capability-index,a900145c33732b7630999946882463edc01bec654bfda2ca46b64943f4fd2e01,eee24ef91c209829c18cde1deb267dae08d236cf1c96b6de8bd64cc387a15b1e,183ef5bbd8fee695abe6f30f22f05eb823283ef3c76b75a2020e2cdf06f5cc37,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,144.97,144.97,exact,canonical-source,profile_f87c6a70e804aea75a8cc3f0,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,b5247cf05973005e61469b5d09302860947e52fbdca5e1748b921bd916bcbdff,52c868039bc53144de246dd3ca96fef81778eee7e6f393420be9e79874d4f165,40bdf6a667cd4ddcd96a1096d2b6ae84a7f744cd7afeeb17b8991346d4f7aad4,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,11.656346,11.7,exact,canonical-source,profile_2b05f8827831fc092f12f06f,reported,unknown,,unknown,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,arena-text,Arena Text,preference,f37c27828b2411878edfcaa5f15c7ed9d1d4b4d03943485a394a41087ff84615,8173e7338fd095d8af40a8be6526adb319620a6b0e26ed5c5878ab5b741007c6,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1455.6556795012555,1456,exact,canonical-source,profile_bb75ca61c28fd70310b7ca1d,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,livebench,LiveBench,general,f54aece242f5a0cdad4a01fe4a9fe2266879e60cb313c0dc66376909d352c293,911666b5a072bd9578610804e5389e232c10328537b47c55174931588443da54,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,63.9369880952381,63.94,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,46608ea3ed41999e70a813961117b0330e306e8dc8a225d70afc2de0b83af97f,8be0be9ac29fd521d7066d62759f2d8f485e41573243cfb233fd19f1ee8cb136,e755250c6a771e29007daeed1c9bda129d6a0ba401149b0194df6c01c47516f0,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,18.813716,18.8,exact,canonical-source,profile_2b05f8827831fc092f12f06f,reported,unknown,,unknown,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,simpleqa-verified,SimpleQA Verified,factuality,181b50a814e7dc7f9d86514074e89947dd00afbb64cce2706ade0f862ddaa0f7,21ca1d3e894829c9d31b3be87b5a932962f299a76c251de4a40e69e1784a845f,82f77751ebf4e0e1bdb747f60520ab845ac1968d26418478e88298d4c1f1b07d,release_ccacd6f679c9142b534364b4,task-v7,verified,accuracy,percent,true,33.081190115985876,33.1,other-supplemental,supplemental-source,profile_995ea331927f1b2b53b092cc,reported,unknown,,unknown,,false,2026-07-23T02:56:47.235286Z,fresh,101,1,A,1,A,0,A+,official-benchmark-platform,0,official-benchmark-platform,simpleqa-verified,SimpleQA Verified,https://www.kaggle.com/api/v1/benchmarks/deepmind/simpleqa-verified/leaderboard
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,gpqa-diamond,GPQA Diamond,science,d5b2e674ad4a3ac6809c0acf42099634fc3371a2fd6b2acf4748ea1e8549b65e,f859c2ac56fa949b22ba6fe50f5eff44cd0ec030c71dd6aff5267c3a6c3ee986,bfd0934f0fdd6356dc91ac83d1c0c6e5d27c68469b95969f85826652ce5fa3c7,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,83.838,83.84,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,mmlu-pro,MMLU-Pro,knowledge,7545fe023e680cd7fbfc22170f6c3d1048905f1ebcce0a4fab3114a1f9039c36,215b545ea3f12af12a5c92c33e3b3fa8bbaab3d46f519d22c8829bb106ac5bda,d89924958304a9eebafbfb87a22248d405abd0873e58816a47bce84c5cba35fe,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,85.836,85.8,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,33e7c34eb631a8cf0aa484d840a374ebbbffe868ea937e996d8a0e20a5fd0968,2c03e51daac3816c0dd5741a68c904ae14335644ac2a11c7a53b8a9a1948da65,f7647352e260863b823a88cb30a738a5bb8bc76b558091650db4b364c51f5230,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,71.11111111111111,71.1,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-08-06T22:35:36.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,frontiermath,FrontierMath T1–3,mathematics,fc31335e017339769cbb8397ad6342d813532b875e3b15849b40697b99df3bc9,9d062ea5eefc098613ca9363f3bb672e378a17a8a951413f37b615946edbd30f,7243a9334d237e65559a5eccb07320f679a3321dc0cd08c2f02302f8aa71cddf,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,25.964912280701753,26.0,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-08-02T01:20:17.000Z,fresh,91,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,arc-agi-2,ARC-AGI-2,abstraction,32c376a7c16a467b1bbb156108d4b066a54c70affae32ce5e62b9a6874b12a9c,b34665bb70682468cf75eab864dc037ea50fb694c0d5478e04b77ff4a850e62f,7b05fe75cdb7f2a11f099e9b0f6748abdaa81040fcc26e106ae9ac77eea282ce,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,10.277777777777779,10,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,frontiermath-tier-4,FrontierMath T4,mathematics,869a1ea366740bf7cb03ebdfe6255ddb8317cad2fcfbe2c7f4e0e56ddb2d68c0,b72bb3e649c29cbac1fabeb16dbd98495bbf1ac44b2dd0e81708e60ba976895d,34ca0cd997d4eb6e78ce323fa6abe83669f340371db64e91ba8daa4505c9d13f,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,0.0,0.0,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-08-02T01:20:17.000Z,fresh,91,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,mmmu-pro,MMMU-Pro,multimodal,652362f44bf83a77eb349cfbc5a25d1fe51c6eb9f54f47d8f0444951840cb231,26d8978253e18f2a86d1589f576663bd7666fd0cba79c17f75c166450a5ff25e,05654b4e3ebd298ec818a7e1cfb6b89a15bca8e14b648030b13070f15d3d6e1a,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,83.584,83.6,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,critpt,CritPt,science,5cf27c4367106e72bcf4e24f7f5a418479b4837f5b5ebefca43f2ef576548dbd,5baf0b4e9406349a3d35084b09b1946ddb5318f78141f6c31699b906f6126307,9b6a1108021470160fff32d4495f470eb3e033d9a36ac83f2e28d9d3ff4b8cf7,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,0.0,0.0,exact,canonical-source,profile_2b05f8827831fc092f12f06f,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,chess-puzzles,Chess Puzzles,games,a51571488cb309793cef80e307ff6f3a7f3f88a84800f279dd40f1ee7a49e528,7b247c0101de440729abb346cea39cc820c526cca3c48ea71dd3d36838df5049,12087d4d9063d130367faa3690a4c04c4d8d32c83607d85273de2251a4ffd6bd,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,16.0,16.0,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-10-13T13:26:52.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,arc-agi-1,ARC-AGI-1,abstraction,2274e344b0e7cd972ee465aa0248050a38c6804c63310fbc367a075166c56579,8f67af6320136f380c211ef965edcbb277402bd16367b18a5da551774fb8bcc4,a393ad5a030262b708816637dea83dd77cc9a95eeaa58236c8ac0eaf5ce4da59,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,53.5,53.5,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,swe-bench-verified,SWE-bench Verified,coding-agent,fd4238e933416eb9b77b07f09d54994cae91cf71ebe116fbad15ce81658ed7a2,87e54aa74fa9fc63817933306ab9a21515cfabfbfb6759b7a393b47750cee0c8,aa9e5fe5595971340220fbe22e04bc88af750823bc50e2ef1279d67b99b763b5,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,75.0,75.0,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,livecodebench,LiveCodeBench,coding,51f31cfaeaf7d98a9ab2b87d6a3054db565801f407a1d7201d589d84f2e8fd77,1a2a1deb2e6b68f2fe66f4a45444deceea72148719acb9945cc3660a50d312f6,ca9fe315645eda54af70f86577254ffb901dff50419cbe4356e6470170c36926,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,79.014,79.0,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,scicode,SciCode,coding,d6531c184215f77cb855bc350f8c852174b91e7335f4e3563c1d57e0021d8c87,8da0a21f26c70355ac3df220ea2fee6a632f6052308dbd1620459e88813aaa69,5c18565f82b5f82a99e1759dba12b495cc395a3909714746dfb5548dfff967a2,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,41.3194444444444,41.3,exact,canonical-source,profile_5ebede4cb3502b6a8f56583f,reported,unknown,,unknown,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,ale-bench,ALE-Bench,coding,ad4372e6efd0fced1e7aaaab6a79634d626f546caaf8ad27b06890d3ab3a478c,b1d950b527047898224bfbe0e07e690db31222c68a0267fdfadcd17100831e51,348938f38e06c3caecddd546cc73a045febe8ba29b77598e5161c8e6611427f2,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,765.27,765.27,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,vibe-code-bench,Vibe Code Bench,coding-agent,a6f8667f649864b4dd928154687d5ac8d3c294642349cea4d8e3eb7d60f686dd,f55e5d442e0f254b0a80a9e97e3da0202204db86e439f86b52129b51b623f412,b73ab4c87960bb7c86dfe9122c3ff611cb904a916653756e25d59c036870a98d,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,37.161,37.2,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,webdev-arena,Text Arena · Coding,coding,2bb4653b24b55c578958fa7607df3233c4003dd2ee6b941483e911148f1f4ea5,924e4aa65450bd400b72adde232e1890a10448658efc7a10dc7436b6b85d71d0,2d298ea6c3afaa5a6794d24a79f426abd0eb6c0bc9b0191bce3b01e6ac94f245,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1434.0835335612885,1434,exact,canonical-source,profile_a7f1f18e36f8a6a9774d349f,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,programbench,ProgramBench,coding-agent,f21add701696e052df7aa01cc49a452fe40a2b7edc039211620c4c7663a640c3,841e031abb0e83b0b58981b0548c736ab2fd3674fc67eb1d7d6ba87351a7c49b,9abf9eefe0749ebfbdfa6c4c8991b89cab0db509f96c2ad07fb75e3b96163dfa,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,19d8528ef2b2bb69218dce143bf36a28e3a317f411e1569022467c5233ccd020,f574795f4dd0a374f74aa4fe0797e7b931566b5fe4f43fb745c5dff258e097b5,d845ffefe50ab4a3d375a7c3e115002ba8f871c41bce3736426fa2357577f157,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,50.187,50.2,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,d5c467da9776a576f47100b625718c94ae269b0a4bb9cf18f60d16818c79511b,eaebc47cd50ae0a465bc7aa6fc2d04ab5d16cfbb1b4002db6adc0f4ebca8082f,bc2584ee2b1d70a51115d84c9443cb0c62b455314bf6459d923694f68755b74b,release_b7d9ab744ffac0cb856e114d,Terminal-Bench 4.0 (Artificial Analysis),mini-swe-agent-66-tasks-500-steps,pass_at_1,percent,true,1.010101,1.0,other-supplemental,supplemental-source,profile_2b05f8827831fc092f12f06f,reported,unknown,,unknown,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,5debe262168dd44de0fcddba32f81a57be781d2fc75c5b778153ef34ece73ad3,c35eac79b163d4c679c30ddf540615168b0b84ac2e41c2d1875db00328826677,3c201b2860ac1f5c9ad90c3bcf78311507ecb897dfc7b7b46841a6fd00dc5189,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,17.525773,17.5,exact,canonical-source,profile_2b05f8827831fc092f12f06f,reported,unknown,,unknown,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,apex-agents,APEX-Agents,tool-use,2334580f6eab47bd3815c82ece7e8b3c73408cb9c6fe5533a0a5ca18915a7c30,4c91deb89989f49b1ea720879980e9f1accf383709ee1bd16f440f3529a6ff98,d5304fcb71b69961e463378cd326f65f4c89ec8e98dfeacdc7a30bb9ffb09643,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,29.300000000000004,29.3,exact,canonical-source,profile_758dd5bfd286cba45f679ea1,reported,unknown,,unknown,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,486b432ba5a258700ac96e0b008ed6589176c419d57530695605e50a6100e3e2,8db2a11e8103726102be6e05093db24711150a94530fa2c132a7f76a7fe50360,63b033b05ce51610e40cd20bf8cea4a045f2bf35d7db9c6f66429e788187d2aa,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,994.52,995,exact,canonical-source,profile_2b05f8827831fc092f12f06f,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
google/gemini-3-5-flash-lite,Gemini 3.5 Flash Lite,Google,closed,69,weirdml,WeirdML,coding,4ffcf323b94c054588f77379e240a666f9f0ff8b295484025885f542c73264e0,e1aa43f8f7d0882e6ae650522438af572dc13aa697ebfdd4a78e6de06a44d778,e4e6c29684eeb5a121905f301b3c516feb86f2bdf63f3e2b92e356eebc3124b2,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,39.0,39.0,exact,canonical-source,profile_4ffb600ecd36bbcb9902a5ac,high,high,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,2be7f1bda3d055a992860409e0e321bd2bb0206f088dcc2972d401c065b0d932,080845a64d5974db2183dbc5088fc76e61a547925dc866d406fc777cde09b608,3bf450b2dd1feff000934cf551b37b7f81711a3a0a5ea6ed60c74ec7ed05b81e,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,26.727698,26.7,exact,canonical-source,profile_85514b3a25f5a223df6b29bb,reported,unknown,,unknown,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,arena-text,Arena Text,preference,1e0bd8b76b1608902fe84d623edd17df924615246ceea99a8c8d6b2ebec960c9,79fccf891070abbf4d6a53557d63ae09de3645acc87550ecd803faca9406d714,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1447.5919518386747,1448,exact,canonical-source,profile_739d929af0ee1a5fc9f85891,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,04eddba024850f557c65f4277992fa6257f961b280fd458a5bda6f263c823f1f,4d92729d52d5931f2ebe4f1ad83280c3729b2f55f06aaea675a6af53fc8e9039,aea26be14761d30a39ab6b42b63bb46cb2fae5b0fd9b650f34348b2703c4c087,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,35.078777,35.1,exact,canonical-source,profile_85514b3a25f5a223df6b29bb,reported,unknown,,unknown,,false,2026-09-30T18:53:23Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,mmmu-pro,MMMU-Pro,multimodal,7774e0b99c9c62dec712cb62fafe9a63ef95e086b6f3aa5a70e39545942500e3,463f97d62c75a15bc48760067072079ad7f3ac4fdc1c19ee691e95b7702f8b79,cf2ff18e1b5ecba733442cfd8e208da258dc1a2f6499599c3c76451eb940d3ba,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,73.063584,73.1,other-supplemental,supplemental-source,profile_85514b3a25f5a223df6b29bb,reported,unknown,,unknown,,false,2026-09-30T18:53:23Z,fresh,31,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,critpt,CritPt,science,02c446c92dfd371c670fc90cf8e5da5bf16b9e999e82dcf6f3b0e83ace5cbd2e,5ae5f6577c022433e40eedda54c3de746ecb0486628b7c84fcb891498f6506ac,1050d57b8c90ed84f807e77aa6b64b932ec998855a4d549f08517dc3027004c4,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,12.0,12.0,exact,canonical-source,profile_bfb973c8007511c06bf8df90,reported,unknown,,unknown,,false,2026-10-08T22:23:13Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,ioi,IOI,coding,0cb9282fc63e3e90337bef30b1fed29bec111af2dd186bc1efbb8287283e1bda,c4e179d789daf3229b50ea57c4ddc1fa2492a714b2e4d8958b870f36ae24ec95,816d6eae1ebf10635f84a86541436e36d196b04b5c5e2a0a99cd146a7d60f7e2,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,47.667,47.7,exact,canonical-source,profile_776318aa95331442d0fd0253,reported,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,scicode,SciCode,coding,05b1db6d8f615a037140f214b2110fa18dd3a5c1269edbb0f8562c0fe18bce01,4b770f6a6f37209f82eee9fcfae9260a186e55ca8d67527e4699926fd4da4d1a,9e501f129cf1e1080e78fd9cb2d033d7a045901f2cf39b03ca8edd6114573b8a,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,51.273148148148195,51.3,exact,canonical-source,profile_8e4d9e7ee373f6ab1750d0b5,reported,unknown,,unknown,,false,2026-10-08T22:23:12Z,fresh,23,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,ale-bench,ALE-Bench,coding,324a04a93cbbc9576ee03383be7c15183778d436679ea42b9b01a5dcabecca47,8d30b63c729ac2369ac4dd3505ec5edd7f024ec67a92805a1bbd7f18c2a9d06a,5a6b3a426453768deba4d814ffb61d584a908b763ee38409283890dbad9a453e,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1244.2,1244.20,exact,canonical-source,profile_5490b7be6211e893b364b100,reported,unknown,,unknown,,false,2026-10-14T07:48:15Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,vibe-code-bench,Vibe Code Bench,coding-agent,1d1e5c8fcc5f6016d714f32a5314cb28d89659b2f125e286877f66c160f334f2,526682c16d72182ecb799a5694e6a1ca0ece9d90707cdf70d973ced5ae765d44,67ef5e00df3a20560d2e2b5b4a98f178727c194f4d899c4f4b2848b1dc0005fa,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,78.96,79.0,exact,canonical-source,profile_8ec17d78c34548ea7a909552,reported,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,webdev-arena,Text Arena · Coding,coding,63a76c9fe98f80f327034007759c7940cdfa95fe3786ab9e8a7125f34c3409dc,16e94eaf93f04bb9dfd53ef559eb7cca17cc84a41b99ccc1dbc657e5d1b5fc7b,9640fb9738a199ce26e690d635ffc5df763f7c504ad2a0bee69d72f5b54ab48d,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1637.1737027632755,1637,exact,canonical-source,profile_c68b1d0ccaccba0c7b04181c,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,programbench,ProgramBench,coding-agent,485bb0c3bb26790c8e37fc8f784ebf3048296db8d972b442b7ee2bdf18790c78,0867649a356cd073bc02e83859258af19824d0f234848fd0b34a84abca9aeb34,d9f60bc8654ea0e1628e7e3289c8331050f3f8f48b817fd4bba45e839b47f4fd,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.5,0.5,exact,canonical-source,profile_eadf3af797290d105cf76c8e,reported,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,8e2e6e77474abed64a3c1e7cb6e35b7034e7b64f9a90d4364aa3e95f4b03a76e,6884861417ff9e151c07721716944b360c60dfb5b49fe2b6265cc8ae9bfce33c,6fc864e40cebd502d9f799e4542b5511049d094fa47b1498196d1633ec0e87c8,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,76.404,76.4,exact,canonical-source,profile_8e264e308d488f525e186738,reported,unknown,,unknown,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,5db187ccf612b7aa53f51eb657914cc2f3819c21b9e6067333eee2f58ed02790,c5a25741dbbd99b2e7091b76950a28d2584ab5df57b205e53192244665bfad22,99bf38ddec6145befa84c29e0243621e4d84ccbcb32e938073dcd741f245bd25,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,24.242,24.2,exact,canonical-source,profile_38513e9f4527d51f190d8a5e,reported,unknown,,unknown,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,sre-bench,SRE Bench,terminal-agent,4e52ce2bfed51e18b77df8baa7b603d4356041e91afdb65e7e12423d1ef34d98,d0eb2a5d598d9270fd3d2e8a5235592f2406ade2fc78fe2d0f348fd9a724f1e9,aa4e3ed0a0beaa2f06a2c3ca5954433150da479237e154a773904b2ec6ce4efc,release_a9585b1cd27a4b33eda715e1,Vals SRE Bench v1,vals-sre-bench-262-instances,fully_solved_rate,percent,true,4.198,4.2,exact,canonical-source,profile_5c611f9309d05bc666dad304,reported,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-sre-bench,Vals AI · SRE Bench,https://www.vals.ai/benchmarks/srebench
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,arena-agent,LMArena Agent Arena,multi-turn-agent,a47cb7a1aa6570d18ba1e650dc843ffafa55183c7c8c6921eee4c4c7842524bb,b706554cde71f784e0076da8a95c57ad7d4983ccd2ba8fef81a0ab56ae968e05,c80fa6143e1f9af5d878a5781010e2f6af34c92e5532c8c158c4704ac6a00f35,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,0.010018747709004657,0.0100,exact,source-specific,profile_424a25c5919bec60800d439d,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,a4147b7b426c5b38a071d58f03e5613bd076ca3434a1b0ed1932f92931c84460,b91c296b80710f445a63194adfc9592bca1732f15fdc78910d40e09e23fc80ba,e85ee6b2d9029d8f4ce994bea4c51ffbee42fd97dc88b4abdd92ef41bbff0892,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1613.56,1614,exact,canonical-source,profile_85514b3a25f5a223df6b29bb,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
xiaomi/mimo-v2-6-flash,MiMo V2.6 Flash,Xiaomi,open,70,terminal-bench-science-0.1,Terminal-Bench-Science 0.1,terminal-agent,3428f07102013f2d0b878a6ffa109eba5af21d784c603beab21033bd6a320a38,5e38a6ea2a8bed1e649f56a18037e6027ae8f15503cecbad01c52373f2044b78,563dc32a594b288031d313c17cc3020c7b37c9e9534771c7976ddea246877458,release_80cc5f4700134aa978bd968b,Vals Terminal-Bench-Science v0.1,vals-terminal-bench-science-0-1-overall,accuracy,percent,true,4.286,4.3,exact,canonical-source,profile_ab8ab28546650c918ceeafb1,reported,unknown,,unknown,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-science,Vals AI · Terminal-Bench-Science,https://www.vals.ai/benchmarks/terminal-bench-science
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,epoch-eci,Epoch Capabilities Index,capability-index,0e222ae668d3cd72a7f66d6ed4e952b73f896ed897b3209eb4722a8adae22f44,692d5046816576b7f1f1b6748a1f05ff7e52f86ecd35559d8765adab84d2257d,ef07353eef6530a9cf40e4dc0d4acac8027769e0fc1fada569f77caa5c04d9a3,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,147.81,147.81,exact,canonical-source,profile_bb2a46687273d3b0b9692880,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,arena-text,Arena Text,preference,6db2e4c0b75c205cad47948b05c7478f2d2462588cb7df347fdb39e1da3ce82f,52f8f6428ad7edeeb8c21317794d0da63e57c453a542f3716e1520945027dab2,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1443.001075193124,1443,exact,canonical-source,profile_53cc057b7b9f88727dc017ac,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,livebench,LiveBench,general,5ebd0aeca5b826665902c372ab1aa1bb2ac6acba5bd119500c91001a9427c4c6,1300e420057d2c7b8768e18b03574295ffb041c90a8b614d941cb5aea3c8fc91,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,68.90613095238095,68.91,exact,canonical-source,profile_4729b1465639558d924d0a64,reported,unknown,,unknown,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,simpleqa-verified,SimpleQA Verified,factuality,4f15e8d0567ea7a6740989fdeb91c3bc2616ac675aba8b81a207beaa62054948,fb73350cf7df919d666c71ee5c0ff5ef47ebe54ca4f5a344e947e5409b38a162,dae7bde98b5cba1f6aa19f37811b32f4fb6cb69223e0e46edb2d865d8fd12424,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,44.14414414414414,44.1,exact,canonical-source,profile_dd42b471745c5d39f5d66180,reported,unknown,,unknown,,false,2026-08-27T19:34:06.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,gpqa-diamond,GPQA Diamond,science,92824f82351bb62920d946fb658fd3b2ad57e39a8e0c0a509da3527cdc84089e,6a2eed6137d926a61b1c76646dbf8dd7cc821c304399a22755557d0f4486e462,7bde3626bb1ca1b7fd4527a66a8dc0174c2023686d8f492a075ad644ee2536fb,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,87.374,87.37,exact,canonical-source,profile_6fe23e846fb827f588fa93e4,reported,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,mmlu-pro,MMLU-Pro,knowledge,4b236baf1ae6754ef95a1ab755638ea5b86b7ec5f45e69bbc4586167d71bd289,8e7ad5410865aa57765c8158f1ad2051cb4819a8c3647145314217fd1c69d533,b3f97980e3465f273004c3504aa3842a9c9b397360b465e24a2b1808facc81e2,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,87.668,87.7,exact,canonical-source,profile_d5fe971d7fb213c8c9957675,reported,unknown,,unknown,,false,2026-04-10,stale,205,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,6f182d30207aa04692c8094ae0c0ada7aa5f1f061a074554d982d5e10192dd89,77302c95c97a068a8673bcc1c3713e83534866b642ba6244c1d8f170a99e3037,a92fa1db59b2ffe3eecd89b77e8dd6e5b3b50056ab0da8fc1a68ad306cc3c8cc,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,93.33333333333333,93.3,exact,canonical-source,profile_edf2f7cfb47d9da6a872a4c4,reported,unknown,,unknown,,false,2026-08-07T00:16:48.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,frontiermath,FrontierMath T1–3,mathematics,80dcbcc064d1784d33e500ce32588d5783b1ce138bd25c04bf661d5a76bbf74b,a5584439798b8a1826c9b282c7637b55769ae99fa818406dc6d79fe541578cea,fc60dfd114311ececfc1f57f9d9205b49465b7c6848b8e45b6aa556afb02d848,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,38.24561403508772,38.2,exact,canonical-source,profile_776ab91b9e881c9d284f77fb,reported,unknown,,unknown,,false,2026-08-30T22:34:20.000Z,fresh,62,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,mmmu-pro,MMMU-Pro,multimodal,7c0a10aac17377d108c94158e17151590f37b2d4117ac24947235a6a2dbc2aca,8ef56babd01235f7c46ea7a82b111aa764ad6acf39910dfa91aeab7696fdba23,804e392a224d67dcc66657f293d7d91fe50c590d7a0c1e8144a951bab51807ac,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,84.162,84.2,exact,canonical-source,profile_cebb40ff99c3605ea64ee86f,reported,unknown,,unknown,,false,2026-04-10,stale,205,2,B,2,B,1,A,independent-evaluation,4,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,critpt,CritPt,science,8e2dd85344d203d1c8063989831ac0cdf403c1692385ff6eb4578e87fb7953a0,556234f97b445e87578369da9171ef40be5c6908a2003d27adef4fec47d87cbc,e5a3c77a8b1382af9e48713c15079fa8b4cd1a9317f97f925927aa6a2fd6702c,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,2.85714285714286,2.9,exact,canonical-source,profile_5525cd61422623d479a15e5d,reported,unknown,,unknown,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,chess-puzzles,Chess Puzzles,games,ece2c1078453cdc56274fcc3e6e9c03ef193284008b7b983e508fcacb575c7eb,4369eb1beb6453e80c670f381cba1eb9257be19cc1480161bcb09f098226d8e5,625f589d59546247992766c32ec916010e01510c1e75947f19a35a981bde75a8,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,14.000000000000002,14.0,exact,canonical-source,profile_6d51b15144d16df42f8e1846,reported,unknown,,unknown,,false,2026-10-13T13:38:57.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,swe-bench-verified,SWE-bench Verified,coding-agent,80691ce72343ab2759a36d2993051aa17d4ab6a4a10b8fe1b86f4d23af804e26,ce8bb10b8ae09e17d95d828a5ed14a2c0604ac686e73111cfe44fdffb9954ed0,907788d9df3db26e653033826be48c07bfa237e8d0fa10b2aec13ffe70cd459a,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,73.4,73.4,exact,canonical-source,profile_f708ff27ef338fc68c717da0,reported,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,livecodebench,LiveCodeBench,coding,8ccc5f7c0bcdb696533c66a3028da13f1cb906dec51a9bdd896d9177b319902f,cb5352f5174f71fe5b541f45a5e9cf7b0b35b15b868f67ff4f1474d7b96386b3,94896bd28a2a3a213dbfa0a9341561acd80c8f6055acdba6975bd16dff04cafa,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,85.952,86.0,exact,canonical-source,profile_9a3feb733930bc036f1939eb,reported,unknown,,unknown,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,scicode,SciCode,coding,c3436ddf182a85ccfac5881c279fbea0f769190ce7a97c01cf9141fba9cd9990,2dbea69cc8b3041085a2eb03029be1760c9add2c96af6bb9462258ce217d52bd,51e091bd6dcb17b0442880d3b04b67d9a6262a3c333e620a64de850ae5139c31,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,40.7407407407407,40.7,exact,canonical-source,profile_3af555d70a9697e425c2e5eb,reported,unknown,,unknown,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,ale-bench,ALE-Bench,coding,2195b1d82563cdef379e6c73e7b4b7da83b295c1ed9a6d13ec931501a8f898bc,85ed3fdddd49caf8644347352c69a8dd7e324eb391a85a28569b14f42145694a,dc65d046db4bddc92135c312fd5d0df89094887ab9bdc53dbe4190c45af29f24,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,670.15,670.15,exact,canonical-source,profile_8201aa5a717e76400d712e5b,reported,unknown,,unknown,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,vibe-code-bench,Vibe Code Bench,coding-agent,0214f2c9ba91166b78b9c00f895740434661f675bcd6dd562f818c27c71f9a80,e5b0b021aecdfeeb3fdc8b19538648492f9eafeaf80e81173ed0f3650e37614b,34848578537378eabc04b9bda4b996f653c89e78bc3e007cc85dceb41649e004,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,25.565,25.6,exact,canonical-source,profile_54ca3d703000d3359b931ea7,reported,unknown,,unknown,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,webdev-arena,Text Arena · Coding,coding,312aa74f6228324cb45b0a5ad7d87a917d162ef7439379a3409018571e73b6f2,05e1f648fc05fa85e2692cf4950c6f727458609144ba87120d044e519e23c80b,ec99a19a49606a7a2aad51887ce8ced6450b614aa29c7d4fca1f610d443abee8,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1461.6254246112119,1462,exact,canonical-source,profile_7a85759823e37f67a2277eb6,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,programbench,ProgramBench,coding-agent,3622d274b0a66efc84716d2a2d2ad8a4423af56aac286fc525d50245aca22ec4,040ac6afe604e27bdd35400e536e551ea10f6c08c8e15534a19586c86f36692a,7a7622be0fecc552d5290f0faf34b0d5840db914a4ec59f28e683d3e02d037dd,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_7262336c3c21354687157662,reported,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,af10fcf479f2343c3f51d41876fb032af9aeaaffce85956b06ddbf699529788c,fcc667f47fc14aa096c9ed2f826e92dae3dfce6d9090de4b8199e9cf93fe847a,18e2faa06b936154a282d0e5ff9fd9f9464985db78629b560abc8ceedd77e37e,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,53.184,53.2,exact,canonical-source,profile_08482f26f2f92123fd7df7cd,reported,unknown,,unknown,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,vending-bench-2,Vending-Bench 2,multi-turn-agent,03a39a41b4064033c928afb509e30860d1160ffec8c549dde9d55b0f8ae5109d,0d5ab237e59bf326dca3d0de3b005b2b2e55e64c3a67e6198d7bf99c8367b32e,5245d7d556f59316a3b62d871bfdd955706a9c52ece3c9601115f737cc29975e,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,5114.872000000001,5114.87,exact,canonical-source,profile_953234b43d2b7322c944aebe,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
alibaba/qwen3-6-plus,Qwen 3.6 Plus,Alibaba,closed,71,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,4c94fe0c920f269d51a625f07e2dd0ddc9d41b89318f123bfb5a7a718370b57f,da16ef72b9365ff2b94f6a1b3f465fff534501a1e93da9800784e7bb651d81aa,d926e830a31b5e8774d52502ae217973a9811c692b83f29a9da3a7daa9e45093,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,996.87,997,exact,canonical-source,profile_93750ce462b5c95291c9db8b,reported,unknown,,unknown,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
zai/glm-5-1,GLM 5.1,Z.ai,open,72,epoch-eci,Epoch Capabilities Index,capability-index,8d950a932701862deb878dde0071b7fcf0420717d6c61cd2ecffebb5478254c6,33e3b9de58dd53c8dbf2f833273a85732fcee2d16c5de37d2779a91d2a41a08e,d5726cecc847a0e03678d6fc8bfa1fb653963a4361a3a0d9b875e42678de5d0c,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,149.88,149.88,exact,canonical-source,profile_dccc52ab99879471547d0fcb,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
zai/glm-5-1,GLM 5.1,Z.ai,open,72,arena-text,Arena Text,preference,b2e72103d434a16f480636ef875b7a3e23db56561ce45c6e005ea64df6cd5986,3352643e7473ed0b2fbca88b81152c5045bd6a1bef35a96418789626f5597344,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1465.3780546778316,1465,exact,canonical-source,profile_4f6d1e7644946448f8c862a5,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
zai/glm-5-1,GLM 5.1,Z.ai,open,72,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,66ec7986f4c10236c40b538bde7f1b2212821b7bb5fe9064ca96432e688c301c,927de019e81ab3926f263cb59e3050d0e831664ddece7352561c3ce0221d58e1,ee783082ec47e7a371a58d8a675a96b1f52ecfbbc63322f29a55942e98f28297,release_5f3023448c2ae4632fe1a43b,reported-runs,public-multimodal-mixed-run-configs,accuracy,percent,true,25.63,25.63,other-supplemental,supplemental-source,profile_431e0630dbeb9073dc747d67,reported,unknown,,unknown,,false,2026-04-07,stale,208,2,B,2,B,1,A,evals-report-official,0,evals-report-official,evals-report-hle,evals.report · Humanity's Last Exam,https://evals.report/benchmarks/humanitys-last-exam?tab=scores
zai/glm-5-1,GLM 5.1,Z.ai,open,72,simpleqa-verified,SimpleQA Verified,factuality,d701a88aa5f8d57cdc69fa3f4a50042a645deafd3f778068cdff3e74ad53df19,32fd2e92293b85ae3c5bf81fb9cfab0673d105bbe2d3e3658550085e396832db,c82c0b1cfbc31d6ae84f5ace46374d5aec7d3ac6893533ed29941bee1c4deb8e,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,34.0,34.0,exact,canonical-source,profile_0284afb597a0b16d0dfbcb57,reported,unknown,,unknown,,false,2026-08-27T19:36:21.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
zai/glm-5-1,GLM 5.1,Z.ai,open,72,gpqa-diamond,GPQA Diamond,science,a67959861cd1ca9fe16ae9dfa2d9b52938177b796db8f7ae6dbc72a3d4c73c79,6da3d76d82f034f970689072dd0d3001aab6a647c3c53e310869db5d230d741c,0ce04b3140e47969c05d67c732517161fcc6a639bf4aaee03928f17e6054e2e3,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,84.518,84.52,exact,canonical-source,profile_b1fc7c73685381c163b9d8b8,reported,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
zai/glm-5-1,GLM 5.1,Z.ai,open,72,mmlu-pro,MMLU-Pro,knowledge,7d7cd1d8af004a34c292c450779686871f809369293f419a0ad6961ebc503180,3f07c5d6be3b0afb38f44331c7f5b516a9309e4ca2b4af138c72b85bcec910ff,b012d9c93bdb3737331c40dfa9ca7e905f4229670db3d3495886856c07a61823,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.9,86.9,exact,canonical-source,profile_8758124509145d977399bb3b,reported,unknown,,unknown,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
zai/glm-5-1,GLM 5.1,Z.ai,open,72,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,e2a332ceb605e4e4540ba97ab63d6a9c1b271b25d27e7c0c7688663357c1bef2,c2269daa80f6f9075f81a0bf0e7a832a06edd9779eb5581234717f4a1c00f43b,392497471e5edbe75c8a86361b5bd275c44d0a931cbec936fb5c638af6cadaae,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,93.33333333333333,93.3,exact,canonical-source,profile_b4a08efe49d38b527af1c6f5,reported,unknown,,unknown,,false,2026-08-10T17:44:52.000Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
zai/glm-5-1,GLM 5.1,Z.ai,open,72,frontiermath,FrontierMath T1–3,mathematics,e2df9098dc7d7896b003d12d11d570686794c12c7818125a689ef70f3de5253e,2efb077e7bdee66b5351ac2dc407c18029f1c6aace3e11d62bfdbfe869dcb21b,98dcfe9815fb35e05957550b906d55eb847952b2b3400aa5542360e7fec9708d,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,36.84210526315789,36.8,exact,canonical-source,profile_48e96ad44abb27180373f267,reported,unknown,,unknown,,false,2026-08-29T11:06:22.000Z,fresh,63,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
zai/glm-5-1,GLM 5.1,Z.ai,open,72,critpt,CritPt,science,ff42a035ac4eb39c83aef8a3112d7c45d2fdbe454acfdbe1e060ae416279fcb1,9070b2a9d2fa098b64532949e213726c5995d8c7419c39ed22c88eeff7af4481,c001b94d23a5145180ef0494950146d455b821d9c6ac0adb93a4a086462d965d,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,4.57142857142857,4.6,exact,canonical-source,profile_4527a4a9bd084290f3d0c928,reported,unknown,,unknown,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
zai/glm-5-1,GLM 5.1,Z.ai,open,72,simplebench,SimpleBench,reasoning,680688d97c68ec50bff4b30c6a5568b1b0deec3415e81c9ca334700daf42a282,74c2950e6348aba3b8b44117f2894018bec7f8d9bb86a1d1084991eeb749cc77,4e1356ea08402efb38774ffe2ae9e2d6cec6e73c9c6b46736562da724a69ee60,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,55.1,55.1,exact,canonical-source,profile_843141e4bdc30672d535a46d,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
zai/glm-5-1,GLM 5.1,Z.ai,open,72,chess-puzzles,Chess Puzzles,games,e584ae9cfa2cb54916349e932535505dd2a10ab112a9e0f78a34a59f904c45a2,cc01942685bbee342d2836b02daf092a3776b1bca984d4d7d0ac8a8e420da115,badc78142d3907a4565b4334b12c13ce70fa9481bea32404fe76d9811f034a14,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,20.0,20.0,exact,canonical-source,profile_c0691b96512398ad384aacdc,reported,unknown,,unknown,,false,2026-10-14T06:28:37.000Z,fresh,17,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
zai/glm-5-1,GLM 5.1,Z.ai,open,72,swe-bench-verified,SWE-bench Verified,coding-agent,69f0236b001f1fae53f3e58d75fb159c5683f3a756708644ef6eee0ab571bac1,3fc23a0cde86fa1b1f2fbcb57a4033bfdb386c99b64a016c0241a9a0df049b4b,b49e3996435128974d669abd88890cc57aedad5426c26ac2fc6673ee7cec67ee,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,76.4,76.4,exact,canonical-source,profile_ecd426e90f3b64f9c038b350,reported,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
zai/glm-5-1,GLM 5.1,Z.ai,open,72,swe-rebench-v2,SWE-rebench v2,coding-agent,53bad1f8d9084c8f2f6208c2bd80f77cc837084939865ec6099ed7667b143920,28bd29daa57451d9be9db498c142c26167d65448a40340d04b936d5d06775743,adc23987b5b7dec5bcf4893f7ec22af1b131cb7476d633bfc720fba4b5732bb6,release_1f8553cc12d8235844f1a578,rolling-v2,2026-04-15/2026-05-15,resolved_rate,percent,true,46.666666666666664,46.67,other-supplemental,supplemental-post-release-window,profile_b8e62d13cae9686240097b66,reported,unknown,,unknown,,false,2026-05-15T00:00:00Z,fresh,170,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
zai/glm-5-1,GLM 5.1,Z.ai,open,72,livecodebench,LiveCodeBench,coding,6694987c2e071c76dcfbff8d13451e0ea4c153701939c65afd707a1b3aec8459,a87670b99f4e01841ed3b2ca68e398df6b08e1fb9e5167eac8596a90d261fa26,83562bbb61ccde2f666ebdf8b836276a198e6e19fd9841ce8794fee35764629a,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,81.38,81.4,exact,canonical-source,profile_d4071b8077bf3ed5a584e7a0,reported,unknown,,unknown,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
zai/glm-5-1,GLM 5.1,Z.ai,open,72,scicode,SciCode,coding,1479bd3b44632ab8d9222a1adc832c6960faefe76ca9e2a12f52a39cf88678b9,153043d7ea1b7528f0e12f72fa661a4ec65aeb5ad8cb5a123bf455a8320e6a10,255b918a3964d48cf69d2ac2b7c2427870abb9538b96a76deff23a505934f694,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,44.7916666666667,44.8,exact,canonical-source,profile_1a321a729524c98a8dd06766,reported,unknown,,unknown,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
zai/glm-5-1,GLM 5.1,Z.ai,open,72,ale-bench,ALE-Bench,coding,03fa75b21e9af4523e1b49d4338c584f786e63dc4c988787d29006902667ff20,b687f0c978a6d1f5d465afef355789599f6858b6aa88931b6b344dbe3bdea52c,87ad9863a296e4d0ca4c603bd65bf083b55cd7df19331b42c8fbfb054df34ad1,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,887.1,887.10,exact,canonical-source,profile_30b8a704bb9675a95ba357fd,reported,unknown,,unknown,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
zai/glm-5-1,GLM 5.1,Z.ai,open,72,vibe-code-bench,Vibe Code Bench,coding-agent,5e86535df85f52e03d60b09f1abeb5bc10ccc3f4bd4ccd8649c9482bae33319e,1911669198cad1ae029728622f562af431402502925ae0f70b04157250a9b44f,3af1c2b5962b59029fc2521e9dfa4c99c359d2b7bba03a08ec8d1d29619b1534,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,31.456,31.5,exact,canonical-source,profile_a3ad892dee5035b3e24ce86c,reported,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
zai/glm-5-1,GLM 5.1,Z.ai,open,72,webdev-arena,Text Arena · Coding,coding,8f2f978b6a5b4946419a4c39a91386968f26e9f6e539d6b900e070c704872771,4ec65f90e377f68ac798d27234ce93b4976992c9350a679e55989bcdc6ca4c0f,e1a41faa44a628fd84f89995f23648ed3cc92441054140acac72a58e8ac145c8,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1507.854558177699,1508,exact,canonical-source,profile_c3270c0606da8408f8bddd6a,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
zai/glm-5-1,GLM 5.1,Z.ai,open,72,programbench,ProgramBench,coding-agent,f8dfaa0a67941eb5bacd798ab67b5caa2789e74f4a699b60a21d15d46fd2aef8,4741286bb31e5d3cdfc37c9de5ef86e471052b45c2bc80cf9ebd671a51ce1264,e9c329ccb4fff8bfeeee9340118a4196d0c40e57e837ac4c5784d89b4f24d272,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_88158ef15864c64c4d9e67d6,reported,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
zai/glm-5-1,GLM 5.1,Z.ai,open,72,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,b7fd27295742960933c65736e48d963b0b24122fa4fc8425c219e69164ce0984,41e0a01ad13fcc2969c339f2c9d920c1ac1efcd06e4cddb74abef801b9bb2b7a,4f2ab12755aabcb57b111a9dc2c589335440f79b13c6bebabf4e8c5cbfec30b1,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,56.929,56.9,exact,canonical-source,profile_3e84dc8d0f289a1a0746e436,reported,unknown,,unknown,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
zai/glm-5-1,GLM 5.1,Z.ai,open,72,mcp-atlas,MCP-Atlas · Scale,tool-use,5b5f60625af33473118124419473f853b042feb43bce37ef426e39b4b9c060bb,bc19a7edbcb7d404e5f5e32701a24907c5ba33504c91b7a8afd173cec9bb39f9,c2c5b28519e379d6e5a699a7e6d682d665e3e8facd48628a0a51450cb1e4f616,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,75.6,75.6,exact,canonical-source,profile_ad3fab65029bdb5cc6f0ca02,reported,unknown,,unknown,,false,2025-12-17T17:20:04Z,stale,318,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
zai/glm-5-1,GLM 5.1,Z.ai,open,72,skillsbench,SkillsBench v1.1,tool-use,9f892f7e5de52b1dda5e89136236d83cc2ad0e8a37423b0fd422b0e97bf5752e,deb148730c332174819980f9b06a681381a2cf295d78afeafd7e77c680984db8,c2ed2025e534826e5e627446f4493a128269710491b35eb2122abbaf19d56304,release_093b45b8b752b3f08534f6ee,v1.1,official-87-tasks-with-skills,pass_rate,percent,true,58.4,58.4,exact,canonical-source,profile_7b5a3c7d436cb995e1ba2cae,reported,unknown,,unknown,,false,2026-06-16,fresh,138,1,A,1,A,0,A+,official-verified-leaderboard,4,official-verified-leaderboard,skillsbench,SkillsBench v1.1,https://huggingface.co/datasets/benchflow/skillsbench-leaderboard/raw/main/leaderboard/skillsbench/v1.1/official.json
zai/glm-5-1,GLM 5.1,Z.ai,open,72,apex-agents,APEX-Agents,tool-use,723a1729387414f2d2394dfd782945c6e54be553c227d3687d27b9140bfddbb6,22186f9b791ca5ed1608cef5524fb11335a973a155f9dbf813d82dfd991aeb26,06977ea746630ea040a4af35202a813d0c631d9475e8aa985a38f6fe25b0b6c8,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,40.9,40.9,exact,canonical-source,profile_8040d78652f9243c2fd1388b,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
zai/glm-5-1,GLM 5.1,Z.ai,open,72,vending-bench-2,Vending-Bench 2,multi-turn-agent,630be6f28e4957fde711fc8c07b3efee4132f91a671a8eefd8f1de7428cd06fa,2de635465465b016e30308f149eaed7fed4be3c5ae7e9597033eeca74020c2d5,9a9e32b0068f5fb9b0b566ae38d4d1d61625904821eac60cb0d101ea97d1ad67,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,5634.406,5634.41,exact,canonical-source,profile_7ec5cca37db5e5d05e3414c7,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
zai/glm-5-1,GLM 5.1,Z.ai,open,72,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,c9d85bfad973125dbca32e3ca36c809d138b9f95b6aa8ccec97464c65cd80c85,9d2d6cdb025fdbfdb4b399e3f21be8482ea68e22cf7110b6df3ca6cf40d47d73,e690e12dda0a8732479dab4d46d4e4368bbf2ef00f28e3f1c0a716acd2775ad1,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1125.15,1125,exact,canonical-source,profile_41edc6ecadbdb6a4fe4eb5b0,adaptive,adaptive,true,adaptive,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
zai/glm-5-1,GLM 5.1,Z.ai,open,72,weirdml,WeirdML,coding,9b6c4e0b761a5a308cf5fc99d85af06d5661bf6edf8740f57c3e2e4fd3d5cafe,71bb7636474deeded6e50c1604a157c33e6fec47ec4673f52d365915c0ffaa5b,c0b0aa6e7a0042316f3fbcf1b37e0959dbd6fc34c9e041d04a2c348019e00d3f,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,57.10000000000001,57.1,exact,canonical-source,profile_3d8decf657d2d47a2ddf04aa,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,epoch-eci,Epoch Capabilities Index,capability-index,fea31ee2030916f72c1e1a5e2e5be4036edfe3538a000ec778e3c78f65bfb076,f147106b19ad5e81dc4a6a22ea8647ee632b01d32e02428bd6e9c9d922e75d30,f328f08ee95979ac682c4abfdc01b6d601bfa4f68198239a044ebd3b8863a15f,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,150.59,150.59,exact,canonical-source,profile_89fbbe37639d7d4a810fd6ba,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,7e5ce32a6b2c1b48dbf8f66bb1b95f6f549f3f39e2af003f07cff01d2bcd5c70,2c4913503eb4a4fd9488885036f493363a2fcf8c1082470dd19f207ef53f5771,72a3614d7ddb6e869bd77dd10693bd086e6917298dd2233d44e3097391d5d4af,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,26.024995,26.0,exact,canonical-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-29T17:25:13Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,arena-text,Arena Text,preference,1d108cac9a15dd1a570dd7dfe6f5ac2042b9c310878e29f03de511663e05c4d8,6484dde58a8d3eef2bda213d7254c9cdf35f1be132772f35c72df3f891907cf3,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1432.0109534562546,1432,exact,canonical-source,profile_637360ad7de092d772e4aea6,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,simpleqa-verified,SimpleQA Verified,factuality,741853d219db1e23f185d1ef33c633eb7b09cfd40b3f5a2347890d93060a526a,4ea37ff47cf79fbcf43360ece9bb8594e3aed692188339d371f4a3faf0676a6a,71640f6fd6a6993f9d26ad035ff322bc310711b93b8bb32a0c23711884bd4e96,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,20.0,20.0,exact,canonical-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-09T14:12:09.000Z,fresh,22,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,gpqa-diamond,GPQA Diamond,science,ae2cb3f698fc0740e083caea8d98269899be843d37d0afcf1a1f8bd390e18471,c6952e92284fad78c5e5a0f08671ea65c9e9f4fc2dbb6baeba299096c2923834,eafa1188cb8307e531713aad7cabd1576533adc4a375ffe5ab13104266b6879d,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,88.06818181818183,88.1,other-supplemental,supplemental-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-12T21:38:22.000Z,fresh,19,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,1bcb9409713653c59711712ed8277bd1c690afe4293e4b3320eb72c5c8046016,58a124551c0d68a31993881937c43a244e5a5502e1a88246176cccfd2e895810,b73e84034c057e1964e23734c4f295ee374c8ea2c88b3daffb33c99109e09075,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,98.33333333333333,98.3,exact,canonical-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-12T21:38:35.000Z,fresh,19,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,frontiermath,FrontierMath T1–3,mathematics,07d414fd4e707515284889de005d78a73708bc723b72e8c4ecb48eae0977a0e7,d9893cd3fbd25d7a7c99f476d59797f1338a50dad3b83c446ace7db191f465bf,78c058eba9fd821332e9a4a7b28f1d8c50a9a62f08ba0d2089cc3f363f89533d,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,58.59649122807018,58.6,exact,canonical-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-12T21:38:35.000Z,fresh,19,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,critpt,CritPt,science,98b90b7b1ac94aa7d4d5eb63bb8bf8b6455dc6b6d257a9b7b6d3e29679c7c009,7fa4a6cc114e0c18030d11c5efbda9f31f526a76c3c492f3f0c4b7dd0b3aea04,0bdc837c4147ddba40158e898fb66d212bd9f88c687f807dbd1e8cfb555c4ec6,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,10.0,10.0,exact,canonical-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-29T17:25:13Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,chess-puzzles,Chess Puzzles,games,721079f32e93e686993718ef160a17b8641a197656e03822a1db0970b4246881,95dc5a961a6acf3652e828cc88a2972c2b55ea99d5034edf13cc90c0cb75578e,b8c53b8298bec11a963684a3cf22323500bf4bf8174dccc074053a250bfa8dbb,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,33.0,33.0,exact,canonical-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-15T22:52:59.000Z,fresh,16,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,frontiercode,FrontierCode 1.1 · Main,coding-agent,f1c409beb64f5cfcc23fde951f5eb22cd8815262dba5f5f313a9e9fe6bd58f7c,9de29ea87e8c75d59f6fb523b5718f1b97ba8497e7db7d8aba7494b842956d25,f7096b525ea21ac7ffd4c2fc717fbe3301c3b8a9436a008a9fd62b2700ae40df,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,39.43,39.4,exact,canonical-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-14T07:48:11Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,ioi,IOI,coding,34b13a89ca118794ea24311a32ca981408cea845039b88c02bcaaa38ed915853,2c078e27d3a6d376c12464f86e85627ddf42cf77b01b2ff3df4b45772d74e586,484705d2b4288049f4fc41628f0cf70c35e5c4d49b2c206029d7ff36b7d081c6,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,45.278,45.3,exact,canonical-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,scicode,SciCode,coding,c9452c458a5a6d5cc947f517fb62a711a1d8111c062bb87acc6ab45bd33622ac,608e2ddbab2c365d396cf604974f1d2ebe329b3e1ef0c56c0e54d2ba9d265a89,b1976e7b09f83d3a68ac82cc1e91e0ea6ec9185f818ae7076c6e7b9788600687,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,54.1666666666667,54.2,exact,canonical-source,profile_e7a36457424cc102e321f255,reported,unknown,,unknown,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,ale-bench,ALE-Bench,coding,907c24379dfe69d8a9a45c6ff6b69df9b61f79ee88a06e59db5c29136f699a66,9d99760531a4e57095490296fe4b3ec5c8ced0eae495da900d7fe0097638fb74,b434bbb3739ad1d8f985a32ab4b38c418a3f26d047352ec15b12da16f2eccb3a,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1213.15,1213.15,exact,canonical-source,profile_1db8ed00113ebd56b6eeee71,reported,unknown,,unknown,,false,2026-10-14T07:48:15Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,vibe-code-bench,Vibe Code Bench,coding-agent,8647c6388a0fb44738afbd40d0f0e45bd8687487420731fa10b69f8d50f4f646,682c1065bf78036563ae5469a8df1d8af960fcf82ec709878691977b5aab426f,6a21e284b909db3b1ba9694bd1a9e6546fc7199b35f791b899a7981a0a8e943a,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,78.403,78.4,exact,canonical-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,webdev-arena,Text Arena · Coding,coding,bb8365699b930544710d1bb8ae6c20dd5b2d865b06cd831fd2d72c306c463687,e8f7c940e05da8ee21e3d5d97121850bc56081c6f163a181d7e51b3fc42c30cd,d4b74c2755b7c79e15f7441580a179646cd8e60590f682265f512bb043b56080,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1544.3103315032786,1544,exact,canonical-source,profile_f6a46b105d68760c1113c3fd,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,weirdml-v3,WeirdML v3,coding-agent,f209950cae598fe475c81aa56831ada218000def757e51b6c3f84728fef29189,1efa74646bde0785625c29d9d3a10cc34f0e2fc6cba26c0ba992b926821b0bf1,df59242b561d30dcd9bd87c81bb11c559ed43c2096482a2244fe965b9aa2f05c,release_d03eb71d3ee6d672c5bf0012,WeirdML v3 current leaderboard,weirdml-v3-11-agentic-tasks,token_efficiency_weighted_score,percent,true,4.445937272727273,4.4,exact,canonical-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-21T23:21:03Z,fresh,10,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml-v3,Epoch AI mirror · WeirdML v3,https://epoch.ai/benchmarks/weirdml-v3
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,7b89d9060ab7d274f7105947e4399f72c0184bc99b3ca92b2ebb2d547cde08d4,8b17f1bd4ab3d381d032c6d7cc69b2e439df6548dbcde4ac09afa2c4b102e37d,65af58ecd810507f92b01f8880c00b6441494dbadbb4c76a53cad2ac93e44978,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,27.778,27.8,exact,canonical-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,apex-agents,APEX-Agents,tool-use,c549ae40adda49247b560f77c906e198b346c8998ad274a0dfd7f703a53f20e0,0db61f09ccca218bd6b746536fd827d20db8a5f9c2a823466789614518091dec,5353b965ce0464d23483a370879260efd1c355ee36b6d809cd75053fd03d0386,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,51.6,51.6,exact,canonical-source,profile_fe87d445715ee13ac9066924,high,high,true,fixed_effort,,false,2026-10-28T21:25:02Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
mistral/mistral-large-4,Mistral Large 4,Mistral AI,closed,73,arena-agent,LMArena Agent Arena,multi-turn-agent,7cdf1a898f11ed2925838820071c1ce3981ac3c0cc3754d490abdca503e3d605,3dbdeddb1789f4d4bbd5b78a9eaae84ed97de07b7ee9c793f1f722e6819af246,870170970b6c7d3adc20c896a113851ec636c3334c80cac9804d9777d55fa332,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.060236922606071455,'-0.0602,exact,source-specific,profile_f8e780749bceef18fb954bc5,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
openai/gpt-5-5-instant,GPT 5.5 Instant,OpenAI,closed,74,epoch-eci,Epoch Capabilities Index,capability-index,1b4a6fa3753cfc8f422a20783e9cee2009894421a4c120aa76ec16795312d080,735576d5f48455f14289c939e443e5970d2c42300f4029ac57aab4541dd2fb8e,4e8d563a3d60bf54d8a07428016e82b4e8d2cf4d778b5d3d1b5e513ce8981914,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,142.49,142.49,exact,canonical-source,profile_68f21033bf83c6f8fbdc9a46,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-5-instant,GPT 5.5 Instant,OpenAI,closed,74,arena-text,Arena Text,preference,a8322998eea0c8018ebae58a08b281197183b9ca6d7641348efa8c5ef49dd1d9,ac287559c91840868e0eabe3ba5f28046adf05e9603aee2b11306598a915ce3d,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1473.1665868835428,1473,exact,canonical-source,profile_330cfe6b7fe94f55d14565b7,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-5-5-instant,GPT 5.5 Instant,OpenAI,closed,74,gpqa-diamond,GPQA Diamond,science,a575577c2f683043e42d9c0dadaafcec84030ac21344ce946ec5a94943ed30c2,a226ff0abc0c9dbb2d7a66934fd3669cf8dbb60bcf6ff0f607f013dcd950817d,86d538da57fca9d7c91572c668d7e41e63f4c09d12c1080a6443217a5173c20c,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,82.51262626262627,82.5,other-supplemental,supplemental-source,profile_e642c9d37f1af0ef22ec3a14,reported,unknown,,unknown,,false,2026-08-02T02:46:37.000Z,fresh,91,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
openai/gpt-5-5-instant,GPT 5.5 Instant,OpenAI,closed,74,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,5725450316f3a2758ea2a962d31caef7a2e25137174529c770502978d8b0d47c,091d3b1f751ed927d44d7b1def28c085b9be3cdc4342c243db9892a63a98493a,bd58235a562a986d5655b9fa3d5e4e24e5185b5caa9dee91090a4bb9727082a1,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,68.05555555555556,68.1,exact,canonical-source,profile_13b94eb3aaf17b8576fe19ef,reported,unknown,,unknown,,false,2026-08-02T02:46:37.000Z,fresh,91,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-5-5-instant,GPT 5.5 Instant,OpenAI,closed,74,frontiermath,FrontierMath T1–3,mathematics,c47821518b62bbf17a4a300d42b01ea13ce748e8be5294c969992443f32ccabe,864943e5920d3dde3c696b02e678608aeca62bca31aec83938fe033895681898,d88f5a819f66f4d19b051f0d673bb006c3c4fb4295841672338401037a42a31d,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,26.31578947368421,26.3,exact,canonical-source,profile_bd5c6fce5440443fd1b5a873,reported,unknown,,unknown,,false,2026-08-02T03:23:36.000Z,fresh,91,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-5-5-instant,GPT 5.5 Instant,OpenAI,closed,74,frontiermath-tier-4,FrontierMath T4,mathematics,7525d0d53c5f8eb01c50a7778d83ac4db9c109c2c5bb1ceb22445dde4a450e01,c2ff5af20fc9ac53ff4703ac1a38e23b7bccb38c4c6ef7179279b7b534370640,3f9de5475de6ffee0c5290dd7ce2901ff8918a8cf5990eb39af29fc63d5fd6e3,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,2.4390243902439024,2.4,exact,canonical-source,profile_4e913c37f9e4dc8418e49c97,reported,unknown,,unknown,,false,2026-08-02T03:23:36.000Z,fresh,91,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-5-5-instant,GPT 5.5 Instant,OpenAI,closed,74,critpt,CritPt,science,b1d2d62db1f0482100c7091ae412c7d5b335461b384aec4e99e70ababd23904d,0e10ebd0ad29f9a6b4aa1b1770a8a6b7964ee3eb6e425b0754039dfe00496582,90739739f322573cf0934f54aa88a344a4e4bc9723803153be1bae1ff77537e6,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,0.0,0.0,exact,canonical-source,profile_002ac6ea591ffd737ee3511b,reported,unknown,,unknown,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-5-5-instant,GPT 5.5 Instant,OpenAI,closed,74,scicode,SciCode,coding,54deff0b38f6d2079b25c0524ea0dc594c3cd6b76b0904ef8da1f016986a83f9,c3edc34c876595e11537d352b2691f4c146be0d546c16c79a04ceac516f16e4c,d7550c39386d0afd190a112feab5a43e14e1e141f83a7ea3a2ce06073420e723,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,52.546296296296305,52.5,exact,canonical-source,profile_0e35a90242db3ebf58471e83,reported,unknown,,unknown,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,epoch-eci,Epoch Capabilities Index,capability-index,4aaddb07f945a1e3f549d7d90092de845d6380fdb02ae2a7ba81a2032cf6d517,a71047768f88637b57cce014258a8c0bc662837717df1ad03977ae4b841f8892,0b8859368d7f1e58273fe53363cf70532bde2c845127c74c49351561dd558c26,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,147.11,147.11,exact,canonical-source,profile_e1a173ba85474b81dd96d886,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,arena-text,Arena Text,preference,adbe62f47d1fa97011a896c094efebe4804da53943a3f4dfa15ebdf4bfda50be,804c3eb0ce49003c6218653f5d6fb7e82c5393823ceee7469112f27c7f9e6558,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1455.6645763163579,1456,exact,canonical-source,profile_d371d7d5e0e9099eb888b168,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,960bd119403589c91a595638a9e0e2598df9a1517ac1fd38d926899b2043b92b,dec78fb32c956d9452986860f5f511acdac52f778060d512cf40081c0362e35a,47241df9da3a012c78b147fc8290f4a7b8b00a7dbd528fd5149c65419be8495e,release_5b5563cd48bd10c552a97181,HLE · Qwen comparison run,vendor-card-protocol-undisclosed,accuracy,percent,true,34.7,34.7,other-supplemental,supplemental-source,profile_e292d3c46b34a34fb650826a,reported,unknown,,unknown,,false,2026-08-14,fresh,79,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-8-27b-card,Qwen3.8-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.8-27B
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,7d87a55512e1cbf8d1fb1c9a05964c8950fab8822f709ada66baf7e7aa5fddbc,e8a8ce6a0ae935d0132ad4646d6915e2923331e79cfa4b22905c6eb15f607ba8,9a1d3939d25507da6fe362b6321d1987c1e52dbebdec771d3a536e0d28035732,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,35.634847,35.6,exact,canonical-source,profile_9bb3bf508162eaffa0f3e019,reported,unknown,,unknown,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,gpqa-diamond,GPQA Diamond,science,ff5afe58e124214134aa3ca6afa85fdbab4fb6ba41ea6fa3b584973124938636,675a8cea96514341f94e531b88be9a4d58c0563e2e2acf5ddd9f05d471eb2b13,6da44b0ffa3849ebd85733410cc4bd90109fa36e5cd86f6b89fe10710b3cb3e9,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,90.0,90.0,other-supplemental,supplemental-source,profile_9bb3bf508162eaffa0f3e019,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,91de644b35cab3c2d6a1ae3005dee0aadc723a7b48a8d496b48bbb5a327f87d1,3d5cd8a8db31ba38d7033ee2ef4588b9f0968eba1b77e3a4aa4884937c9ae463,c11db80caa95e906de1780744c86ce97eea2a244063e2f0d567b927e1b89146d,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,93.33333333333333,93.3,exact,canonical-source,profile_9fbc57d50cc30579ef9b8901,reported,unknown,,unknown,,false,2026-08-07T00:02:57.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,frontiermath,FrontierMath T1–3,mathematics,de80b63df64c00f7c027742ecdf9647da0377bdedd08e4ad1427cd8e95f35c2c,0a129c4b1b1f156ca6304bc753948b64c43d806942841e14800443cf64dcaac5,089d994dba8fbbb9a94adeceadac997c7a787574abb88833801d8debec0ea2ea,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,34.385964912280706,34.4,exact,canonical-source,profile_6823403052eb3e76b6245e53,off,off,false,none,,false,2026-08-29T11:07:26.000Z,fresh,63,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,mmmu-pro,MMMU-Pro,multimodal,ea72f44255017aec8f021db36a12e664c3af1f0c44665d9e835a7a77d2d1c3bb,63f1f87963095c9d356f59120a5f92cf7747a3adee519d867d035dac18d8db0b,f93809cb1601fb231715a1087a0f48409f3053a93ff6588860a4cdbfbd0cd653,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,80.462428,80.5,other-supplemental,supplemental-source,profile_9bb3bf508162eaffa0f3e019,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,critpt,CritPt,science,905642a876ba77ce829c929e1838a54bbd99a1e530927ca3c1eea5520d1d1253,a9694b1edee589e53d411f6b1e1e1ab359b358ccb874f5045e2b1dc9657cffa0,36bb2fca4bda91f25f07762b9299161b7c6d16724b0441a070a7833378cad47c,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,9.142857,9.1,exact,canonical-source,profile_9bb3bf508162eaffa0f3e019,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,deepswe,DeepSWE,coding-agent,04c6230f42bf5eeb21096cc35d3653cbc80dc6f63dbe5b4d251b4d22cc409245,ff3f043b1fc70ba355e54f57fc599fe01c8a4d46b8dbcd9e845c073e14db80d4,148322bba9aa3bf90b760832066a5d74c8fcb793a80517a108761e22a6184493,release_0f94a5c34a87744f357d4816,DeepSWE 1.1 · Qwen comparison run,qwen-vendor-comparison,resolved_rate,percent,true,14.2,14.2,other-supplemental,supplemental-source,profile_e292d3c46b34a34fb650826a,reported,unknown,,unknown,,false,2026-08-14,fresh,79,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-8-27b-card,Qwen3.8-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.8-27B
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,frontiercode,FrontierCode 1.1 · Main,coding-agent,2be691c5a540d669dc655ec36caa5218baa78c9639bbaeda2405aa51b005cae0,fe763f9594fc928b24b12205a2fa8e254babe8ae1baa6528b9af42874d0076c2,6503e291c9d7c606598f28e53137bd9e936aecfbc73279fcbd53c05a90ffbc1f,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,10.2,10.2,exact,canonical-source,profile_6823403052eb3e76b6245e53,off,off,false,none,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,scicode,SciCode,coding,9a3a15b4b9bd1ebff198c4c1e39e1be0acf2805f6c9a450e074f1ff3cd38fec8,ab7f87cdc6d98f47b7d218fb49d2ffc0ffd77a4be82f6de2dd50d611e5f88959,1ce3dab97d66b987eadb8844d61f963f5b913af59fd9d397acf55723fc781a87,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,46.064814814814795,46.1,exact,canonical-source,profile_480536e6d05172cf6b264394,reported,unknown,,unknown,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,vibe-code-bench,Vibe Code Bench,coding-agent,a783e59a5fddb32b643d8f818035108fb6395f78215b4453578761b9b5ad0a45,9abe8065d30e888b98de1687c93757629176491038684b4b3745887b82f908a2,e0e882b0ab35fdffb55e468beacc03755573f89707e73a0c80aaeb6aae20733e,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,46.391,46.4,exact,canonical-source,profile_f76a19fb1cd57c1a8d315778,reported,unknown,,unknown,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,2bc8e9da90aecb61f4ff868e78a50d82b6f23b5d7fa8f15196e1948930f2bc5e,0f971e94e5c5e5e5cb542f5024634434b0be1316744aa9050fa8a45ea4171bb8,ac185b03ab001eae7f5132118ef2c4a7313adbd2950c4f51516f792c1cdfc1bd,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,52.809,52.8,exact,canonical-source,profile_a276c3b003828a116e9aecd3,reported,unknown,,unknown,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,587608a517425c07c55c8851893ad010b81ad062abbab39078b4ca9e6cd7be44,d434e0a869a8d76eda5546d399f8361bd4413e97ccb8bce2d7aff23132cc859d,7be448a07df209b1aa0aa5481a894480f413697d05279f0575f1dec13ee36064,release_b7d9ab744ffac0cb856e114d,Terminal-Bench 4.0 (Artificial Analysis),mini-swe-agent-66-tasks-500-steps,pass_at_1,percent,true,1.010101,1.0,other-supplemental,supplemental-source,profile_9bb3bf508162eaffa0f3e019,reported,unknown,,unknown,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,7cacbaf31aa24c35a6a642290761d31baa1b2efc2c445ffd1962265f37350c6f,0f42e09782c1a462fd01d2a99bbd3d386d88c873182de4faa14311beb9357ff0,1d2aded4db688e734f1df56f297ba2ee088657007600375a37ac6d3fbb4d7682,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,17.525773,17.5,exact,canonical-source,profile_9bb3bf508162eaffa0f3e019,reported,unknown,,unknown,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,arena-agent,LMArena Agent Arena,multi-turn-agent,338126d5df2db3e179ace7b83b90d1370c903c7a11fdae82154daf7bb4583d5c,81924903f5e921de5f8f94dcc7cf29f08e45bb259c2f368d932975351511b339,85af7362c973244df4f06de3900d5d0eb524a91f230df99f3103437aaba240e1,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.07368242078467599,'-0.0737,exact,source-specific,profile_73a5d0395f902cb3229076c8,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,81ab6c4bc698b774599bbdde4e323607eabc524af05feeb11fd42fc4b46af6f2,becc7990aeea94c564187e5bd8c96882f43f639057af3b156a068bf244be5046,60d632058a17a8edd1e7e153f58ee1e9d007adabe41421ebf8b73ee106bd891a,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,774.8,775,exact,canonical-source,profile_9bb3bf508162eaffa0f3e019,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
alibaba/qwen3-7-plus,Qwen 3.7 Plus,Alibaba,closed,75,osworld-2,OSWorld 2.0,computer-use-agent,f6fbfb1d0f305a48fc085f1e7aff84e90a5fd031b792235f1b64eb165836724f,fda3dcab3f13e511899df1218677c78d201ece39bfd1d7d347855a0873d48807,6148a7636d450b51ea486e87b225d1c3a61dcb0e7064ecec4c1bbcf6458770ee,release_15c5525fe39f33bc181dfb69,OSWorld 2.0 current leaderboard,binary-accuracy-500-step-budget,binary_accuracy,percent,true,2.8,2.8,exact,canonical-source,profile_d4e18270aaf1ff5ea39c0e8b,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:27:06Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-osworld-2,Epoch AI mirror · OSWorld 2.0,https://epoch.ai/benchmarks/osworld-2
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,epoch-eci,Epoch Capabilities Index,capability-index,b043f361852dd0a6f2628aff0415b8c185f364920c285f542602c8021965c3d5,7410e557e9fd52dfdf44e420d748ff646b8ae3fc339e6e2237f64666059cd769,9d153826daea7706daf0e976628c6cb0d3bc6ae4277455d4f0386b86785301a6,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,146.14,146.14,exact,canonical-source,profile_c58cd8aea3620eea4e04e0b2,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,arena-text,Arena Text,preference,b29f11101ff98f6a523f2a2aef712828ddd4a5730105ec5f8df87245194c37b5,e3baa678b37e481d7abefac54911144be5d55310caa02cd2ffad117b523c4c90,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1401.1879805288152,1401,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,livebench,LiveBench,general,f1880b24c6df4d78d0082b4f8cda1b1605f2c9c20f83cda842992841b813ab14,5b78052bd5a6cc1a6beba5cb663dbb0f13fa5290967f89673714c08c64435383,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,69.57636904761905,69.58,exact,canonical-source,profile_0bdad69e8278c95ccb05d676,xhigh,xhigh,true,fixed_effort,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,simpleqa-verified,SimpleQA Verified,factuality,feb65bca9e3644603b0d1dd802c04d67b3a7d52040fed21145d0e2def65700db,d111f5e5ef96312bfcd04b4a5bd268bca9c360097f2b705e5ea4d1a3a55f8c83,40c85021eb5d68a8360d0cf162dc8111f4c27837683acd2c7671760bb97b1290,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,11.700000000000001,11.7,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-08-27T19:28:50.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,gpqa-diamond,GPQA Diamond,science,cdee10adcc02edc1e10e77ab668e120bd0b71fa0f972e8eb4ece0747089d70ef,1a6a4069437078abff2e3b6bcc79f740c450c3a7e807e4413b763b8a78545e6b,f72e26ad04d2580d59f2a069a20b7b41c4b42747977e5f47fc8fe47a4a27f21b,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,77.526,77.53,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,mmlu-pro,MMLU-Pro,knowledge,640d532f16269fd6058dc28c8001da5a59b3fa56aa2916086e76b25d5d1ac6e8,09073b0edd3295b959cb637db371a4dfdb5aec656877b8bd7b5b9cc90b32cca8,623f1a4152e4efdb2d991f16352ac0fe2f8087a90806b85a6854d6a69e38c31e,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,77.172,77.2,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,f555649ba4aeb1c844035bebe58cfa39d9b0c0de027e0772f9693fd0e0fc9ed7,78354ff77b0bda88f5264c545ff8be27f42a8df59e03133d4f8009fea49c8b5b,533f1ef8c0aa7368cc8ce278ba524911f1b9cdf50e3b24c5b3d290e6cbb3e9ee,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,87.78,87.8,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-04-14T15:08:33.525Z,stale,200,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,frontiermath,FrontierMath T1–3,mathematics,b08c78c0affd578b0d0b52f461d5c6870b6f881c016b4c6c60ad8e6990a921c5,f43f3780db22ace19bb72071bebb55434b8c0dedc1627c68d06919adebc90a8a,93d550bd4f0659ec8d3b3eb5c3b3db2737270435e50ab1c146000095e38607fd,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,44.91228070175438,44.9,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-06-12T16:48:06.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,arc-agi-2,ARC-AGI-2,abstraction,ae2748613ef4f3b8530aab376b93e3c4e306d2b3f789a2edfe3ad1a600d3d0d5,b926d879a1dc3646020f7f476c4d3126ef9801c7e23b59fac460232046f2b08b,01ad944d01ddecc00a82f5739b0aa8928b7e1d102db1bd4cb03edeee749afc04,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,5.6899999999999995,6,exact,canonical-source,profile_0bdad69e8278c95ccb05d676,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,frontiermath-tier-4,FrontierMath T4,mathematics,3777058552c5eba6b0432042b5202b18fcae18c021e988cd8eaeb3daf4d1eca2,11e51bd4855b3049a3eb207c5e93fc485c30b6572856242ce9a16fe0eb0a3884,d893fe105e20ffada7ffbe0fcf3bab7fdadcc1d84a7dc9423bf50cae4db895aa,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,12.195121951219512,12.2,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-06-12T16:48:06.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,mmmu-pro,MMMU-Pro,multimodal,1ed72de538a194379a952580d0d29e39fe192d104737e91854d5f8e3c3170d21,f4f0768948193a4755722c4fc40529e0440897914522e622d4c350333c0f0892,ebe1766a0644371b65ce86b721bcd3eaa876d6dbb47b5e38f43bb8e668fd1379,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,73.584,73.6,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,critpt,CritPt,science,933fecb1b0cf543972b6b46177adf9cb05f47d2458c60269edae00508ca9f3d9,3f9ce7853d3fa9be8436e8c126e796cd618718b3d2be2abcec84e64f53b8fad0,faa9cc272286e3acb5520699808be0fab7320804012d0d12a9cebf6ce040f54c,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,9.253061224489791,9.3,exact,canonical-source,profile_0bdad69e8278c95ccb05d676,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:13Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,chess-puzzles,Chess Puzzles,games,f064e73ce4c0c92db4ecbd59bfc545c902610390783f571e9af2b661a6baed6c,a0455a27f71865ebd6076f79ff9d91725bce68b6bc1711e30105176404e324f1,d8e54fed013e0bd0c52121f1764e7f068ff5e831139177664102a571d32eca74,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,45.0,45.0,exact,canonical-source,profile_0bdad69e8278c95ccb05d676,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T18:22:26.000Z,fresh,17,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,arc-agi-1,ARC-AGI-1,abstraction,e54a988e17fd99fe0e6d4509efe70f9ca0a2a261f64d9334fb08e1852ed22042,7c28fdea7056a624133cd1d3b1f0cfdd3b9cb0f807eb1e36a27898bb91f5a111,6639afb33b297a63129ad1db4133d6f05051363ec0cc4f04a095f163e8058946,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,51.5,51.5,exact,canonical-source,profile_0bdad69e8278c95ccb05d676,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,swe-bench-verified,SWE-bench Verified,coding-agent,63c847df4cc78c57c3d4b5c8a573c545ff87bc1ffca257502fe71ad5d23a6599,9da4bfde58e7cb364ab410113642ce4ca16ea5ab710ea81d3e4709595ff9f9d6,e34f8d70308d221225c6a6b5995d36e4ea38c50c1fcb94081bb19e87afe195bb,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,69.8,69.8,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,livecodebench,LiveCodeBench,coding,1223c0c1ddcaad4645fbbce980461fac4c92fd2782dfc537f96c6235cee3ae6e,5c2cab297461f68f5597e41ab6a3a7c9777ed4babfd0828025cab8099f5c8738,63df3ba2664e1731dfc6a1e5cb7e5005e4878249c8fb64edc6366ad405677264,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,84.009,84.0,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,scicode,SciCode,coding,0cba08a1664d1bd3610b75bee27eb288e6c66817f89298ba4cd1b65fa90aa978,70bcbcb12b672c46058d9ef22a2558bf1133252b287cf77e5b61a1d433d2ee62,774cb81eb6f5c92014a6deb628bdc0c97e393e7196f28fee7da197c267d5d6c7,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,47.2222222222222,47.2,exact,canonical-source,profile_0bdad69e8278c95ccb05d676,xhigh,xhigh,true,fixed_effort,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,ale-bench,ALE-Bench,coding,9e7fa0a79d208a3b11da755e52db9751e31b3f65899f87157f527521f88373e3,6aa3c5885c8dd82617f4b3334d8167817a34894de56d82a53aad84ca1d604b56,5540096dfcc8c3c297c5fa426d9e3ce6369532054938f8a51b1175afd64203a7,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1004.52,1004.52,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,vibe-code-bench,Vibe Code Bench,coding-agent,4f0457b8001941db69a20406e61f16cee0dc2273d0f12de7efebeb86d1355ef7,1271d0c3cf5168a82a5449e11163bf86ee17ee748b64773f25e6149a8aa27e54,bdd04c7612461a8e24170cd07c97e96da8a9b2653374baff5b15904723879c90,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,26.097,26.1,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,71736e1dad4375a3e431f68be081a30ce2640231ce0bdeb42d569e6e9fb68ddb,e395d250532b08a28d1f4022364e03bb0e4046f171ec1fabb5953c05532a3f76,7f8387c87843a1438cd189e32941eea7928dd353e750953be9998bf4afa47e2e,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,41.573,41.6,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,4e9868a18bb92279d99121a1fe9263da2ad06df167376b35548f7a6e9ca17138,7701cffcc546c451b5df8a307afaa08ceb0031f4ef8c129d7591faaa8891200c,cd57b3fbfbf16855b17f40a847e576c77b934f17ff4ff7addd0472cda93ec8dc,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,954.72,955,exact,canonical-source,profile_0bdad69e8278c95ccb05d676,xhigh,xhigh,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-5-4-nano,GPT 5.4 Nano,OpenAI,closed,76,weirdml,WeirdML,coding,263dc56d121746f73279ae75ce123546680d1a82ab2b6a290700332984a7cbc8,3be0a09e2564322cedec1e980f51eaac94d724612268327f285978e75fc782c6,50acdb50a06c35f2d73d3d2a986a6a213b3ef2d53afda42f0f791ddd3fadee20,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,49.23,49.2,exact,canonical-source,profile_5bec2cbd214ab2f1bec419a3,high,high,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
tencent/hy3,HY3,Tencent,open,77,arena-text,Arena Text,preference,cae77396fe9596c576e2b89a0225885c7e63bd6f2d8ba44686098508d0277586,7305984d0004fe014cd7982916eacd297d4fe3381fc2fa36702bc489c3638229,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1454.4774865775755,1454,exact,canonical-source,profile_a509997ff81cf41e490ff500,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
tencent/hy3,HY3,Tencent,open,77,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,c8a127e30c3ab6e3ed54a2c690e500388967fa6b7167692b5449c719d3eab7fc,abb0f27ca6c91deeef5c2def315b862e46ae528af565053f16c9b6ee6f3f9f63,e1f956386a97e6ba42a3c396757902d19f19cdabb10fcc91e345af5c771a2890,release_fa9e908ac04b067eb00182d4,HLE official leaderboard,published-leaderboard,accuracy,percent,true,53.2,53.2,other-supplemental,supplemental-source,profile_f4211279c9136484fa19f66c,reported,unknown,,unknown,,false,2026-08-29T23:52:43Z,fresh,63,1,A,1,A,5,C,self-reported,6,self-reported,hle,Humanity's Last Exam,https://huggingface.co/api/datasets/cais/hle/leaderboard?limit=200
tencent/hy3,HY3,Tencent,open,77,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,8b9c7a1a6bb2ba14131a6cf9f920ccf1f8d8b4f168c49367d0d3d7093f13861b,eaeb1c7497ab461551f02de356846f0fdbaf6b5e8d373a26267cb1f4252cd203,99e293e55aea33c8da4281cb1f0b4953904152d4dcfebc447c94533136e55b89,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,33.456905,33.5,exact,canonical-source,profile_6c679e74ea315691491c34a6,reported,unknown,,unknown,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
tencent/hy3,HY3,Tencent,open,77,gpqa-diamond,GPQA Diamond,science,69b65fbffb5eaf7c17783abf1aa4e9e0cc108f1e7d390e8ff0d0fe36e88a9ec9,a084ac1f910f76478412b3446990065399ae7ed2776457c7399b18771e529ecc,968a1aaa9a52dab88e7556ea31cf218dddcbb7cd59f724c61f9671a804d1e469,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,89.69697,89.7,other-supplemental,supplemental-source,profile_6c679e74ea315691491c34a6,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
tencent/hy3,HY3,Tencent,open,77,critpt,CritPt,science,547dc635535bd401d4db1082711946bd68fe099e02a0fc3147a83f267a0d9677,d8fb57dac56449f1a2902155e77125960197e1ee403cf2b7c8ad501d51130092,557294d53731dd4ee147720dd4cc5020da030224bfdc018e2e139a89a5772a2d,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,4.857143,4.9,exact,canonical-source,profile_6c679e74ea315691491c34a6,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
tencent/hy3,HY3,Tencent,open,77,scicode,SciCode,coding,fa1eb6e10d4b5fd557a7b042c1b251395f2ef06b28be96b69d7115df562140f8,68abdf82eaecc737ea971d984491ec8b183a6f8411fef691db226d4ed0b1f1bb,8069bfb27e3b6e2131beda5c8418ee8f0f925f7d5b1f588604dee47a34436e6b,release_de40c66bd40a7b5483b40f9a,SciCode v1.0.1 (Artificial Analysis),scientific-research-coding-subproblems,subproblem_accuracy,percent,true,48.611111,48.6,other-supplemental,supplemental-source,profile_6c679e74ea315691491c34a6,reported,unknown,,unknown,,false,2026-09-06T00:23:09Z,fresh,56,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
tencent/hy3,HY3,Tencent,open,77,webdev-arena,Text Arena · Coding,coding,679f694b4269aaec3a56c489bf009ee40dfa27e3d94d16f5e771bf15c721c465,59805de91f39259d8974ef6b42012118e45084830d663dc9c8bb17e395d6ef67,3062a6cfc0dac05b72943619088607bff4ea22b102402c9cc8be4543140815eb,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1505.4337074624896,1505,exact,canonical-source,profile_debe9c2cc9379011d1013ab2,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
tencent/hy3,HY3,Tencent,open,77,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,566f3c6b23333229453439381201e086ee26b02992d3011d39a4f49ae64cca1d,b26d81f12c9fa887383fc55cec363b1690f283e40c421973158f1443d8b8d057,50d9f98c8f53af161f65126accb9fe71cf1c7cf6a5e840a31b81beedd97361b6,release_c53c830678318dc1763793c8,Terminal-Bench 2.1 (Artificial Analysis),terminus-2-89-tasks,pass_at_1,percent,true,64.419476,64.4,other-supplemental,supplemental-source,profile_6c679e74ea315691491c34a6,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
tencent/hy3,HY3,Tencent,open,77,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,1b54687bbb293f0f261c5447839a16e0e51e108544df597a4cd55c65e1b46bf9,77a6a8e24a0a32fe2fe5748cf07f8f9c1d69f52776f1106db385d5c91797a476,f2d7519d685297661d499a7fc637adf730f530d3a9dcb0771a8507f792579d99,release_b7d9ab744ffac0cb856e114d,Terminal-Bench 4.0 (Artificial Analysis),mini-swe-agent-66-tasks-500-steps,pass_at_1,percent,true,0.505051,0.5,other-supplemental,supplemental-source,profile_6c679e74ea315691491c34a6,reported,unknown,,unknown,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
tencent/hy3,HY3,Tencent,open,77,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,8f2e8377a8c9d3693792b0a4d92011216f1f1b7ed774a62b3490bc0f9b66b1c1,427ab27e8bb9d597fd52ca33697d7bffe8dda0f956232e9500339f3f3edecb58,901c1c3599875db999a6318b4ba524fec4cb69a2681dfc9635e7d05522a6a9d8,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,22.886598,22.9,exact,canonical-source,profile_6c679e74ea315691491c34a6,reported,unknown,,unknown,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
tencent/hy3,HY3,Tencent,open,77,arena-agent,LMArena Agent Arena,multi-turn-agent,0999662c2a65baf4e3201c041f91d8d892abc3fc1d86ca49815b8f9f2120c9bc,60934a3e4a0ad2c25534453b7e8b6605c7a27d7aa1837a3d49414b7180c0c0cd,df183e60ce4fe8b2852c7212f45ae1ff2aa3f20ca770862225146185ac43e1b0,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.07319108163660436,'-0.0732,exact,source-specific,profile_98914b89929791d91bf8e5de,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
tencent/hy3,HY3,Tencent,open,77,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,34224114512140765deaa0c42e860ff66abffd4ccf5168ce198c56214ffe80b7,bef0ad728b0b9eb59dcd01897a8bdba14947150e1599d5f613ec27f5b2b238da,804c8ce66603cd95625f80c49ab297e8a54461e9369f479c96b81c73ea192dae,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1070.64,1071,exact,canonical-source,profile_6c679e74ea315691491c34a6,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
xai/grok-4-1,Grok 4.1,xAI,closed,78,arena-text,Arena Text,preference,66e13ad60058def31fea513b1e4270be3fcc7b7983d1592cc4712d19b2f1c408,57f371335360a4f161c0d9e39361cecafa33d588bd6fbab701d79eb1f93c468b,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1465.0468418320115,1465,exact,canonical-source,profile_308a2315a1ffdf31f407f589,adaptive,adaptive,true,adaptive,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
xai/grok-4-1,Grok 4.1,xAI,closed,78,webdev-arena,Text Arena · Coding,coding,0adbb9be2843771fc31bb99824befd28ea01c139fd40627102b8c1ccd04fd160,c0ed80ce77e0329eab86818623dfdce67e2e616fa0ecb29c2bc6c1304cd4c40c,e9e202e65e7b57dcfcd00a5ec44dc3dd0764c8ee1c04c0c8c5dfa2259d312a92,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1214.468784837361,1214,exact,canonical-source,profile_308a2315a1ffdf31f407f589,adaptive,adaptive,true,adaptive,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-32b,Qwen 3 32B,Alibaba,open,79,epoch-eci,Epoch Capabilities Index,capability-index,a315889280da0859208840dd8dc55844a632bafc6a99fb6e0c1e84d49e7e0fd7,b6296f25c8e2cbb60e834e554791c0ba4dc37949e3b4fcebb5149798ea75413c,bd447d1b8a6d17ef80da168963c315dbe54affa185e70019a1908f20f73a53c1,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,138.48,138.48,exact,canonical-source,profile_cebf1706df3e3613a3a126de,off,off,false,none,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-32b,Qwen 3 32B,Alibaba,open,79,arena-text,Arena Text,preference,af1d6a67e7531c0f8c7470ce59b87e458718fc7eecdf26edf64adc657a6b2442,0611d496790f1811621a1034cc1151d90f61fb1ea545dda26c9fec7e004e8680,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1346.8185340713344,1347,exact,canonical-source,profile_3aa37c0df95a457752b0f86c,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-32b,Qwen 3 32B,Alibaba,open,79,gpqa-diamond,GPQA Diamond,science,c578cdea3d58140408d5d4da790a56540deb3d3cfe0545d76625472161466e01,691c23e9102bd1e86c1a6cdebd12a8fce26c0d2c254c7801b98a09a44a45fbe6,9f415d65db681547f574fafbfd1adb7f8ea722f85b42ef092c6e4a25fd9cb7a9,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,54.10353535353535,54.1,other-supplemental,supplemental-source,profile_cebf1706df3e3613a3a126de,off,off,false,none,,false,2026-08-30T22:45:08.000Z,fresh,62,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
alibaba/qwen3-32b,Qwen 3 32B,Alibaba,open,79,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,a56b56716bcf2f4555dd9e7fc7d0a954bd1a530c656633433041c1317041d760,c5f89d278b050168600020b469a199b0e07f3955621040c08d98d8b5cd43b204,7b3ac10f47270801e3a35e6dace9a96bafc344848517ae97dda76973b235815b,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,23.055555555555557,23.1,exact,canonical-source,profile_cebf1706df3e3613a3a126de,off,off,false,none,,false,2026-08-30T22:45:19.000Z,fresh,62,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-32b,Qwen 3 32B,Alibaba,open,79,critpt,CritPt,science,0f91fae05dba326ada869c485e4fd7a9d8149a8eb01735bc2a523c1de7ef6460,a62f5885bd0cac2b69539b5650d1a634651ee4f51e6e6c5db78909d6b3a16084,eac59e785a258c2e38d55b2af57e244a9a7b33dcf9cfaa45c3270de9367d03bc,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,0.28571428571428603,0.3,exact,canonical-source,profile_5eeb27f9b68bb2dd1366ce55,reported,unknown,,unknown,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
alibaba/qwen3-32b,Qwen 3 32B,Alibaba,open,79,ifbench,IFBench,instruction-following,c87ee84344739cf66d96a8f54cfe116d10f02e915868c57822fde2fcabfa4af1,0f220e63f7c9a937537f26f094eefde7fb7f7a57bc8630e1cca5b98681770f95,ee7e2333dcebcd23fdc8d1d794d58c4e202e42da23d768519a4735d0c3f880ca,release_4adf348ac2f76539ed7ccb10,swallow-evaluation-instruct,ifbench-test-300-prompts-344-instructions,inst_level_strict_acc,percent,true,29.4,29.4,exact,canonical-source,profile_d3418e26019f2a26b71ed3db,adaptive,adaptive,true,adaptive,,false,2026-09-30T18:53:19Z,fresh,31,2,B,2,B,0,A+,official-leaderboard,6,official-leaderboard,swallow-llm-leaderboard,Swallow LLM Leaderboard,https://raw.githubusercontent.com/swallow-llm/leaderboard/c58a13a852eba061d639af7caf4ddbdff33b36e3/swallow_leaderboard_post_en.js
alibaba/qwen3-32b,Qwen 3 32B,Alibaba,open,79,scicode,SciCode,coding,5dc15a58f66a0e5ec2623c0cb36cea07fabf18ed9c6a1568b006fc195e1055d9,ad7fd25da6f6bdd1182b7f6884f481b87c67249d551f11f9cbc7b1d81babc840,a8a54d7bb5464445029ee47794e5942fd4724f3060920e9fe4a6a64e74e5cfd0,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,35.4166666666667,35.4,exact,canonical-source,profile_7b558a43320ab043f3b5f3e3,reported,unknown,,unknown,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-32b,Qwen 3 32B,Alibaba,open,79,bfcl-v4,Berkeley Function Calling Leaderboard V4,tool-use,cd33793f9ebd5eb7f1920add92d5603dfea50457d9123964e7e5b97d8d22bee7,3a2a5707980d70897715555f206b9c658905ff52e75014ffe7230cd45ed5e559,6a9c99bccbca11748bdf70a2bdd6eb864726c94bf4a029f84ab95898fe96d192,release_7d2560723dba9d549cd10b25,v4-ede5081a24bc,overall,overall_accuracy,percent,true,48.71,48.71,exact,canonical-source,profile_85ab9abe2438a5cb1fc4cac5,reported,unknown,,unknown,,false,2026-08-28T22:03:18Z,fresh,64,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,bfcl-v4,BFCL V4,https://gorilla.cs.berkeley.edu/leaderboard.html
alibaba/qwen3-32b,Qwen 3 32B,Alibaba,open,79,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,72df5610eaacb1b218bf979923218a4143d4bf405b65e31f39b6c054e276eb68,b23a652c87c3dee026bb49fa8ded929583b9002e3866494b3ff4121d1a1cac0b,c4a2978edafe5419f8b06f382adfed858b2b8d8592b93c058d010bc999d0ad30,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,40.84,41,exact,canonical-source,profile_d3418e26019f2a26b71ed3db,adaptive,adaptive,true,adaptive,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,arena-text,Arena Text,preference,238903a13c04278fc44b1c6190ef3920f97034a6c23ce7a32be718915fb058b7,cd6a6ffaaea84dfb29e580f17e56d812dd882f91d45310a6cddb8fde90e8f2a8,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1408.8059895114475,1409,exact,canonical-source,profile_e7d27865cc833f004dcca283,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,aa2dc90a9c9803881d44c1b50e12eb93581ecff4c046d127ebc0056bdc37de1e,8f70acf791c9f0e0ca3cce26e0c9aa8213dd72869af99935664bd2fc013beeff,5193c99f2c93e1107d46b911094f08dc7f18c6366de909d53f5f747daef3985a,release_2f3c6e21d0f3616fdb78f627,HLE · Qwen vendor run,full-multimodal-cot-no-tools,accuracy,percent,true,24.3,24.3,other-supplemental,supplemental-source,profile_f5f4005af8cee78d3b3a8a61,reported,unknown,,unknown,,false,2026-02-24,stale,250,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-27b-card,Qwen3.5-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.5-27B
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,gpqa-diamond,GPQA Diamond,science,90735b1194831c86757ca422ccf5424507f9e5630c1b91d28279b2b73b685146,2a4db942edad015be6422bd3339808797367937820040ec2457a325b9b3d6352,1ff9ae2ef8d730817a6d761b03392d3f8a0e65ea332abf727e3e7f33f3bf7ac6,release_beac9b3474dc330fed0ebc77,GPQA Diamond · Qwen vendor run,diamond-vendor-protocol-undisclosed,accuracy,percent,true,85.5,85.5,other-supplemental,supplemental-source,profile_f5f4005af8cee78d3b3a8a61,reported,unknown,,unknown,,false,2026-02-24,stale,250,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-27b-card,Qwen3.5-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.5-27B
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,mmlu-pro,MMLU-Pro,knowledge,4b1a0029e4c8a18e268217ed3357c4445e5b77badd12b3419dacf9f5775144bd,95819c8f70ac3e6071dec87e6d9ced7a537bbe2f8ed5ec3275fddf3faefe5799,70c7bbff02e177f65267ba3c33be7b6f38096fbaac03c9cc970c7210519303fa,release_95dc04609d027072fa229d3b,MMLU-Pro · Qwen vendor run,vendor-card-protocol-undisclosed,accuracy,percent,true,86.1,86.1,other-supplemental,supplemental-source,profile_acca955d8107d22c457bcb69,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,mmmu-pro,MMMU-Pro,multimodal,f8a0eab458435dd1ccea8a8a95b5116cb156dc04d239b28dd7404d9b45c3ecf4,fa14cd2bbd4c2bf6348145a31d98fac7273764f4cbbfbfec27f4993ae38d7f68,9225c08e521da63df44b0ad31230f50489ffbee930c8fb4aaf2f0f8dbf6fce3f,release_1dbdadbee92bb7f366d1025b,MMMU-Pro official leaderboard,standard,accuracy,percent,true,75.0,75.0,other-supplemental,supplemental-source,profile_acc078689edb6302f6ab810f,reported,unknown,,unknown,,false,2026-04-28,stale,187,1,A,1,A,5,C,self-reported,0,self-reported,mmmu-pro,MMMU-Pro,https://huggingface.co/api/datasets/MMMU/MMMU_Pro/leaderboard
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,ifbench,IFBench,instruction-following,364a58e2266ed75054278deaf8f86b1c610d9ae7cea42a038eb22903f5e917ca,0787a798a6bd1c7f7632699f20bd8d9b6332cdcf8cf1433473fa15196b9d1d9e,9b012eb36f1c2f4cf17e351131da2df24841c3192da533ea9c19d25ddab91dc8,release_4adf348ac2f76539ed7ccb10,swallow-evaluation-instruct,ifbench-test-300-prompts-344-instructions,inst_level_strict_acc,percent,true,64.5,64.5,exact,canonical-source,profile_de190b968e73c379f7673b4a,adaptive,adaptive,true,adaptive,,false,2026-09-30T18:53:19Z,fresh,31,2,B,2,B,0,A+,official-leaderboard,6,official-leaderboard,swallow-llm-leaderboard,Swallow LLM Leaderboard,https://raw.githubusercontent.com/swallow-llm/leaderboard/c58a13a852eba061d639af7caf4ddbdff33b36e3/swallow_leaderboard_post_en.js
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,ifeval,IFEval,instruction-following,0f60a9e8e68f08820766291b50b9518c5af029b98299a55e1948489249a8077c,ad0b1c0431eb2137759034f5a08551f176b3e89a51a66e8fda903018206a14eb,e9e17b8e0074098f32daf83a5b38ae12858d3337cf27da604f333e45a9c74efb,release_eaa2f0262bb0264860cb2c8e,IFEval,qwen-vendor-reported,accuracy,percent,true,95.0,95.0,other-supplemental,source-specific,profile_f5f4005af8cee78d3b3a8a61,reported,unknown,,unknown,,false,2026-02-24,stale,250,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-27b-card,Qwen3.5-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.5-27B
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,swe-bench-verified,SWE-bench Verified,coding-agent,508366e715fd26527334c243cb1c9d1856db291f10619131db2dd0ff3c3ae8b8,bc9c93323c34b402a644ae0f3990f4a838e51c6cbf88e2f8be9773f05b0ae477,dad3fccd5f7c7052561a2997a748ad0d5aa39550e8421c973d8656bf054626e7,release_9162beb27813d52a46a20a9f,SWE-bench Verified · Qwen vendor protocol,qwen-internal-agent-run,resolved_rate,percent,true,75.0,75.0,other-supplemental,supplemental-source,profile_acca955d8107d22c457bcb69,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,swe-rebench-v2,SWE-rebench v2,coding-agent,5581ce3391010b25d03a445cf11f0cf35ff41327c8188d317f59bb10db49ad59,7382af3cd1986618c6ac7bf6028f899e6656b4886399aad60256fe182f2457a3,8fe98e6ef0f26358cd9945f72ca92647ad82c74b572ec206946223c4d5374ded,release_7449132b5f078c71f23c26b5,rolling-v2,2026-02-15/2026-03-01,resolved_rate,percent,true,54.81481481481482,54.81,other-supplemental,supplemental-prior-window,profile_a806723737ab8f1e6f695e6f,reported,unknown,,unknown,,false,2026-03-01T00:00:00Z,stale,245,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,livecodebench,LiveCodeBench,coding,0848261135ac11f532113ccb5d135ef04f91e250deb1ed16917089f7f9d62430,fd7def26cab509f4c29b7e65b1d40d9eafed6a385cda7450ced51f753bbd948a,95d681e3f443f8d5d265d19ae8a8ce03ce12f86cdcded5df0926c1f3f815adf1,release_b530ca4da413885c54f52baa,v6,qwen-vendor-reported,pass@1,percent,true,80.7,80.7,other-supplemental,supplemental-source,profile_f5f4005af8cee78d3b3a8a61,reported,unknown,,unknown,,false,2026-02-24,stale,250,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-27b-card,Qwen3.5-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.5-27B
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,ale-bench,ALE-Bench,coding,62669d48b84e7eedc5ab6ae508a7709e5a2fb54da28a647694f6fa2294374c94,366d79a456cfa56c4b5a234914e8b7329d369e0f3af82929285495980ea9de5a,11a3f9a7201ffbf5ce021e153c092dbee9b3c852f72843e74bf3a6f6832cc58c,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,349.45,349.45,exact,canonical-source,profile_d51b4d24a4d78655be5fdb73,reported,unknown,,unknown,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,webdev-arena,Text Arena · Coding,coding,8315f6b4046cac03bc78891fa386f35062b54c70576da6809c0d9c066fc6f466,9edd6a9ad44cd23672ce2b18057adbd391c2df741413a80c54ec09ca55d8969f,e93b6280a139cab1759cd7e13e0e3e6a8c8d02ac2163ff73134941530c431cb7,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1358.3092755979856,1358,exact,canonical-source,profile_d8f6ad970b239a16860c4cb8,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,skillsbench,SkillsBench v1.1,tool-use,14efd825387128108cdc67e04c9ca2bf2bb4a9545f39549fef3d6f6ce0c51d6c,9fc927ca1b0f86703221415ca3933b9017d9c039297831285860f5a1b4913174,1ef4df8bbd26617f6a97c4748bab4ad0b89cb62c839ee783c7a5b25a8ac69428,release_5af9fe5a9d5e275323798c8b,SkillsBench · Qwen 78-task self-contained subset,self-contained-78-no-api-subset,pass_rate,percent,true,27.2,27.2,other-supplemental,supplemental-source,profile_acca955d8107d22c457bcb69,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,bfcl-v4,Berkeley Function Calling Leaderboard V4,tool-use,81b4ebffb3d2731b02c754f4c014f149f13ca4ef67fc13f096dc738a4530ad27,7c59d77b1784d000e4aa71ca70d868f07aa757e7e037d317225c8ab8b7fbd3b1,9aca4b26ad4c1fb4b38873d2a78f94206d521bb77b8fd1c6904a2c439362580d,release_6b9456463dae72f95fadf6b3,BFCL V4 · Qwen vendor run,vendor-card-protocol-undisclosed,overall_accuracy,percent,true,68.5,68.5,other-supplemental,supplemental-source,profile_f5f4005af8cee78d3b3a8a61,reported,unknown,,unknown,,false,2026-02-24,stale,250,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-27b-card,Qwen3.5-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.5-27B
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,vending-bench-2,Vending-Bench 2,multi-turn-agent,43733dbd9f95bfcc9da0c80d4ec45c08625326e07a3e5f2558c4bb96b6ce00f3,a7306297d7550483ac8814a7406b5aa33c660dcdf32fda27868b7ef44c88f84d,6c1f16e150ac8721678431423fc4af65a8198ef9c08fa3300f1696b29819e7c9,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,201.97599999999994,201.98,exact,canonical-source,profile_7f48d5054d4b2dc26d1818ca,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,529e6c579b42eab22051cd0fc44e159ee11b9b89abe1b206d8498e589d6bab93,3d45b9e0852085f6c506238cac53c43f1ea267ab7296f6e5abe6b9880db62feb,ac033ae84d9e8c3792065e58a44a15f3d80b846134d957b52b2b43ded2458ebb,release_c6e5abafbb7bffc0975932de,TAU3-Bench · Qwen vendor run,qwen-official-user-model-bm25,pass_rate,percent,true,68.4,68.4,other-supplemental,supplemental-source,profile_acca955d8107d22c457bcb69,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-5-27b,Qwen 3.5 27B,Alibaba,open,80,weirdml,WeirdML,coding,83c4ac11c658be15f3d1fdf91bdaeb07479aaef2321cd23130b7928a68e4818f,fbc6dcdd8931bf48b2591c9dd6e4050bc0a72e182714eb347792c8e00b4d9110,d3567b16c7cd49d2fa47f9c8fb940d89aebd07f44bfc96ba13c8315e6ba5b8b9,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,39.53,39.5,exact,canonical-source,profile_9393d6a5cff56c2e2f1c57a5,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,epoch-eci,Epoch Capabilities Index,capability-index,e53be9e7c63976426c4245a70fee7306d948834fe23423ec2cf3f214c0b5b897,8051e6581e535377996a546e29abd5defb5d83781015211a62b286c6946dabaa,4f72bc1c1d8ee33cce785057e32ed8d163986dc9da68ff5581c177263ca5529b,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,142.56,142.56,exact,canonical-source,profile_8038866fb4362de4cdb57daf,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,arena-text,Arena Text,preference,97be34cedbc840d4e721ea03b4c3837e247030a5dd1ec87723f5ba8d9141fbda,8c07d4ff5ee3d4cecb88cc26914c3d327c85a45a7a38a11c250163f712c484b1,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1393.8308645238358,1394,exact,canonical-source,profile_7bb891886f6b9f82e595d537,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,171536dbc27ea4f24a2de988e75a17ccf2194b216ef1ecf4d6a9b649af6e1373,cc316e55ab5d1340ddbd43bdbaeaf712dc551414a11ee5b85c4af871634ee7d6,b49926e35945bdfc4bf990d42caf170ca771797b88a16f10251ffb64499db7de,release_2f3c6e21d0f3616fdb78f627,HLE · Qwen vendor run,full-multimodal-cot-no-tools,accuracy,percent,true,22.4,22.4,other-supplemental,supplemental-source,profile_acf8ff0a882897edb172285e,reported,unknown,,unknown,,false,2026-02-24,stale,250,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-35b-a3b-card,Qwen3.5-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.5-35B-A3B
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,8b19002343d59f5170dcf6102013a817e2bd92ce52a211cb2debaf20fa09e22a,3ca6d672d608f8242ce99f58ef7c4e7c28db27d74857a7e2d2839b51e39d6909,c82f11f93ed09ee059ff29283f4a431b1a4d7e73b955c89638baeca630283a3f,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,13.438369,13.4,exact,canonical-source,profile_08198280c56f559e82197c5d,off,off,false,none,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,gpqa-diamond,GPQA Diamond,science,70b8b395c9bca22bfaf5c6d41778c1c0262516621e53ae34b8184f23bc1e651e,ab8c732c2cb942be11260bf0c77f94cdbb39b3ca9ead81c3688a43075fdee3c8,b73e5cd99b7c054fc9be875f4120b9f11e2bae8cc6e3ec141337a49314ca1bfd,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,81.919192,81.9,other-supplemental,supplemental-source,profile_08198280c56f559e82197c5d,off,off,false,none,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,mmlu-pro,MMLU-Pro,knowledge,b4c0ba1cdf51a27be1e0b529016fcd72f3682b596b5c039e60a3167eb288b08f,0f5d4241e89874f8a37a8a560aebce700d8b24e146f66a4887a677bea0b699ef,70c7bbff02e177f65267ba3c33be7b6f38096fbaac03c9cc970c7210519303fa,release_95dc04609d027072fa229d3b,MMLU-Pro · Qwen vendor run,vendor-card-protocol-undisclosed,accuracy,percent,true,85.3,85.3,other-supplemental,supplemental-source,profile_a489b097f55632aa33d8a051,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,c0c5d7d90220ff77ed08b70921760931eeb63ca407a9494f4dc86949cb5034ab,cf2353eb0490029b74386fc2b701efde22553dfdf975fb75a5d2dd5d68922310,91930fa35f725ad3c356dbd7a26c502c4a1cd60c9d733a134b49eb2cf966042e,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,70.0,70.0,exact,canonical-source,profile_08198280c56f559e82197c5d,off,off,false,none,,false,2026-08-30T22:45:56.000Z,fresh,62,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,mmmu-pro,MMMU-Pro,multimodal,b0fcb6ba75e4e7a20827ad9b91196294953357dad6c2340a92b9f6f4d12d094e,55170197b63105bc97e67f4d77dc0a3f977f968eeb7a3b247fd2ea181ca9beeb,f5aae5ac6ff3c3614cd499cc0dc6857e5b4b834398f1722f1b9ecd67fa0f1ef0,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,69.190751,69.2,other-supplemental,supplemental-source,profile_08198280c56f559e82197c5d,off,off,false,none,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,critpt,CritPt,science,7aa70417927a98bd80a0ac504310bc4ff8008a50add71fcb85a5d3b1b6dad1c2,998a7c4772fcf54ea828c67b94d351b21b1a004bb297fe5eec6686deba04512c,5883adcd5901ca67361c21b62e568dafd4443eeaad296150e4b6a0027d4276ec,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,0.571428571428571,0.6,exact,canonical-source,profile_08198280c56f559e82197c5d,off,off,false,none,,false,2026-09-02T21:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,ifbench,IFBench,instruction-following,eca7d6e060781481802685d019969481f5e9a6ded322c4f48bd9c379a2cb490a,877ddbf051b5bcab60fa8f7770661a20dea8bfc844c555f5c0071e33cf73b5e2,637455e9814ebdb49cbf54e163208050c724c6ab8f9e5de8f6d8bcd522d407e0,release_4adf348ac2f76539ed7ccb10,swallow-evaluation-instruct,ifbench-test-300-prompts-344-instructions,inst_level_strict_acc,percent,true,59.9,59.9,exact,canonical-source,profile_f3e955ec7b7b358b2d54c536,adaptive,adaptive,true,adaptive,,false,2026-09-30T18:53:19Z,fresh,31,2,B,2,B,0,A+,official-leaderboard,6,official-leaderboard,swallow-llm-leaderboard,Swallow LLM Leaderboard,https://raw.githubusercontent.com/swallow-llm/leaderboard/c58a13a852eba061d639af7caf4ddbdff33b36e3/swallow_leaderboard_post_en.js
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,ifeval,IFEval,instruction-following,9acd3fb07b5b7882ec632c448c9b55d601504cc1242ba9ea16104334b021827f,f93a5f4172e3a754eb1c169b80f3846b83e4860f8578764e449d162e948239ef,75c65dff8a4ca072f7779ee663a82f005dc599d80742e995c92e9efe54e12762,release_eaa2f0262bb0264860cb2c8e,IFEval,qwen-vendor-reported,accuracy,percent,true,91.9,91.9,other-supplemental,source-specific,profile_acf8ff0a882897edb172285e,reported,unknown,,unknown,,false,2026-02-24,stale,250,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-35b-a3b-card,Qwen3.5-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.5-35B-A3B
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,swe-bench-verified,SWE-bench Verified,coding-agent,31a81390ecd810045cdffee48ed7e139c4c8244a463f2acba0952d929c8f2fb8,b91ad4b09950831548895f062638ccf94225cc8fcb4af4646b49154793c68906,dad3fccd5f7c7052561a2997a748ad0d5aa39550e8421c973d8656bf054626e7,release_9162beb27813d52a46a20a9f,SWE-bench Verified · Qwen vendor protocol,qwen-internal-agent-run,resolved_rate,percent,true,70.0,70.0,other-supplemental,supplemental-source,profile_a489b097f55632aa33d8a051,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,swe-rebench-v2,SWE-rebench v2,coding-agent,a00999461d5a15b2c4520d00b039f525824db79177232cf2feab5f7fc9d7b05c,d222bdb3496ac12279d1b3ab91a8ffa1a6b2ed382df032656260aa96da7694cb,6e5a0da2f64d15a75563cdfb237ee200ce210a778081a94c786af397623a1536,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,17.117117117117115,17.12,exact,canonical-source,profile_be91adc28b2c36400b8ba1d2,reported,unknown,,unknown,,false,2026-07-01T00:00:00Z,fresh,123,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,livecodebench,LiveCodeBench,coding,58cf9274dd79179aadedb6d26afe7a4de1155417f03ad69f5b7a8715badf18d0,cc1229feb272d3d3078a3cf81e8c6c7dcfe54db5f95098084c9c40be82424d2d,db4e6883454c44a81391e2be412d53f1f000436285ebff88bf47b93dfaaf4dae,release_b530ca4da413885c54f52baa,v6,qwen-vendor-reported,pass@1,percent,true,74.6,74.6,other-supplemental,supplemental-source,profile_acf8ff0a882897edb172285e,reported,unknown,,unknown,,false,2026-02-24,stale,250,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-35b-a3b-card,Qwen3.5-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.5-35B-A3B
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,scicode,SciCode,coding,b6959f75892eb9c90c9a42614a2d015bc5f107edfb4b849d3e9572f522db3e97,30d06fff619f7110d2e123ee855ff152d741937bba31d080f7c2e0a8d87c6532,1a198886d4fc358ff800abfc9a5c92e2bc8dee83167d73c119f59429dd9b55bf,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,29.2824074074074,29.3,exact,canonical-source,profile_08198280c56f559e82197c5d,off,off,false,none,,false,2026-09-02T21:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,webdev-arena,Text Arena · Coding,coding,f779127ce143ee784c219be10be47fee20fe16613c27f22ad7c17e24bad34646,b801c18bd8ed26e4d2b24955fb052d9e3b13e050e3a94310928ddac7db10e2a9,54eb3afb7281c0b80a19a2e50b6b9c81cae95f8e45834f75c1d2452ecd57614d,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1254.1583882182388,1254,exact,canonical-source,profile_629f0be1378c8068b93135c5,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,507bd7ad9ac3547fa5ea4e89958cfc350e23db21026c18c997145828114d4094,e44eac0e9b0094abcec8e3cf11840e92cd0c5297f012db7c5b1e65179e7a2b6e,f2b361ff5b17541ab507cb160241fcafa9a6674485d1d00f853a48c327a53570,release_c53c830678318dc1763793c8,Terminal-Bench 2.1 (Artificial Analysis),terminus-2-89-tasks,pass_at_1,percent,true,40.82397,40.8,other-supplemental,supplemental-source,profile_08198280c56f559e82197c5d,off,off,false,none,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,8a17783e7ac7059066dc1af719a08bb03f28a9aaabdb4cb7cf972b4527f14aaf,ee503dfe9dc12847528d3e6c09025b18de8b273bc5aa31d6b72eff9373ef3a47,86784d5fdedb2bd1b029547fa4a787eeca76d6b99e30a3c187610a4b91db16e2,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,4.948454,4.9,exact,canonical-source,profile_08198280c56f559e82197c5d,off,off,false,none,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,skillsbench,SkillsBench v1.1,tool-use,8d3661f56a87d13740247d4a672dd945c65110093cf50964ec5fd737a16e6507,320dbfb4604193a6eaeef0ad21a11f9eecc81e9dc1b7e148e0f22154b166cfc8,1ef4df8bbd26617f6a97c4748bab4ad0b89cb62c839ee783c7a5b25a8ac69428,release_5af9fe5a9d5e275323798c8b,SkillsBench · Qwen 78-task self-contained subset,self-contained-78-no-api-subset,pass_rate,percent,true,4.4,4.4,other-supplemental,supplemental-source,profile_a489b097f55632aa33d8a051,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,bfcl-v4,Berkeley Function Calling Leaderboard V4,tool-use,ea0af1f6215972d71f3a4b2fd3d790adf4829ea0426ba4cf7d4ce7ee2354ec7c,ec02988d19fe94097e752e7d934900ce8829321a4996eb59f285dad0733ce83f,d2e5e4ded552b5be2636ae22be91934e3af5094a2c843dba75092617f97b0e8a,release_6b9456463dae72f95fadf6b3,BFCL V4 · Qwen vendor run,vendor-card-protocol-undisclosed,overall_accuracy,percent,true,67.3,67.3,other-supplemental,supplemental-source,profile_acf8ff0a882897edb172285e,reported,unknown,,unknown,,false,2026-02-24,stale,250,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-35b-a3b-card,Qwen3.5-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.5-35B-A3B
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,189b23e6bfebe05642e7a75c051fcf70302dc873da828d09ee3433a7456545ad,696ef00a210db809c193e622bb7dbb47f53ced41650c048ee913a29cb261256b,db1e4b69951d9dde73f5e0e2f494836b4fbffdf4e55c6f8f59d9682fa30e8c01,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,616.33,616,exact,canonical-source,profile_08198280c56f559e82197c5d,off,off,false,none,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
alibaba/qwen3-5-35b-a3b,Qwen 3.5 35B-A3B,Alibaba,open,81,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,a67f709ca26801d56ab6e79690408d0807c344540de4f486e3ee4ca2d155f896,ae5081bd124fd991bdcb417a116f31652e02bb443d052ed2c1a9267a1559fa4a,ac033ae84d9e8c3792065e58a44a15f3d80b846134d957b52b2b43ded2458ebb,release_c6e5abafbb7bffc0975932de,TAU3-Bench · Qwen vendor run,qwen-official-user-model-bm25,pass_rate,percent,true,68.9,68.9,other-supplemental,supplemental-source,profile_a489b097f55632aa33d8a051,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,epoch-eci,Epoch Capabilities Index,capability-index,d617d0e4a4a2497a974318701a1710462875781fd9b683a23134b3fa08b87bc8,38b9c8b3576114b2be776fd8b7a523ddbf9557c61bb86196ab5e3472310ff2de,89be2d7e36c5e8e6ab174330a30c774555aafc6665973c1313c1f881b56528a0,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,146.32,146.32,exact,canonical-source,profile_d79e65ecde0b2da0fa5b6c49,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,livebench,LiveBench,general,b0f2da3a93f7a311cb26f2fed44911ae93ec54dbfa12a9653ba9171ea8debaf9,8757d60738d6dece6d2a7676301a88ab2e70a2dd3338207c8573daa2c0305f07,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,64.02713095238094,64.03,exact,canonical-source,profile_9e9091d6396e29d6d8b835ff,reported,unknown,,unknown,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,e78c69da452c465f374145256b5a44a8a2ac79246ec16e38814488e61b1fd28b,7877427bc2c85e0515619a1da87c6a63d1c1a3e24926bb1b2d6c48e7c0de4c70,e1f956386a97e6ba42a3c396757902d19f19cdabb10fcc91e345af5c771a2890,release_fa9e908ac04b067eb00182d4,HLE official leaderboard,published-leaderboard,accuracy,percent,true,24.0,24,other-supplemental,supplemental-source,profile_3570f866437a50d449060a12,reported,unknown,,unknown,,false,2026-08-29T23:52:43Z,fresh,63,1,A,1,A,5,C,self-reported,6,self-reported,hle,Humanity's Last Exam,https://huggingface.co/api/datasets/cais/hle/leaderboard?limit=200
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,ca662e17504badc2405bfdde71290ed5a43c211ca496c306d0c641969b48dc48,47f9dc12d4fcd60444216b4d8052beb99e53fd429a26ed7dad35d82f4131ee65,cf0c5313fdfacffc253255ba0019c8f3863debca16c6843efd7ca675067ba9c0,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,23.076923,23.1,exact,canonical-source,profile_532636b8035b61ccdae69abd,adaptive,adaptive,true,adaptive,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,gpqa-diamond,GPQA Diamond,science,83629bffbd3eef1837f8b4b3f346393530763052e1ed61bffabb03e97f3e3dcc,ec7728c28a745fb0284a471a4587a672bf81cf3c868472d90cdffc7a1fa66c72,fa2fbf8361f9daea6f1967f15ede01589b5889326a0d9e09a2f1c0752c0cbb64,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,84.242424,84.2,other-supplemental,supplemental-source,profile_532636b8035b61ccdae69abd,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,mmlu-pro,MMLU-Pro,knowledge,0a3a26d152b241aa7292684e9295d46dc64da81b6368356d472be114ac9db40f,b91072c0a84c1a6fcf0c4add3474cce51465210c5cfc7de6f30fd732769ddd19,44f56864a490bb7582461036c068400061930b80aee3b634127a43235a3bd532,release_95dc04609d027072fa229d3b,MMLU-Pro · Qwen vendor run,vendor-card-protocol-undisclosed,accuracy,percent,true,86.2,86.2,other-supplemental,supplemental-source,profile_609752b0f7bd8a6341aa27a9,reported,unknown,,unknown,,false,2026-04-22,stale,193,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-27b-card,Qwen3.6-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-27B
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,0718f636a083cc5bd5e319af629e5de137ced38e9543b8639adc560ab9450266,0b967dd3e37d345780f9504fe6c7d8a5417b6965ec67077959f32dc4ccb6e5d7,62c42b96674f98c0a3d367bb21907b85a8fd1fa83ec55da20c0692d255928e09,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,91.11111111111111,91.1,exact,canonical-source,profile_de27cbb86e9c990231514038,reported,unknown,,unknown,,false,2026-08-07T00:12:36.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,frontiermath,FrontierMath T1–3,mathematics,36230e414a754c764a8f6a0d4917bc164a10e4f9ba2087d1216cff2d78ef0fbf,9a67054ea999024b818440d89bc7ce4169b8cc900f4bbf7be2c3c07f0773ccd1,1bed8ed4070c75c3320000f0d35a74716111e019ba019ae7fde7b7195d434cab,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,34.03508771929825,34.0,exact,canonical-source,profile_2cd29242487b8e61b4d8196f,off,off,false,none,,false,2026-08-29T11:06:28.000Z,fresh,63,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,mmmu-pro,MMMU-Pro,multimodal,1b6f115e910e7a1eb328bb1381451099b6657dea065bcb3f07c2e9507728aaa0,f4ee2b33934f53bf5ca91cbc01956e5552390420924e43f0db084c0aa7b94b9c,bcba8686eb1b301d47af1b852f210dbdb831a09d7f5f203fdcf6544309c6a39c,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,74.624277,74.6,other-supplemental,supplemental-source,profile_532636b8035b61ccdae69abd,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,critpt,CritPt,science,92a5b32e1ff83400fd1eba46c40ba069d591ad8cae12df6f109d0f2cd5bb1e40,f7a15a8eb55df5306896d54a3f4a5f2d62b87d11164ddc1d94382c8cbcddfb0d,bf4fee08f497620015d5779ccb4ad100f29c89690690d44b245ef61d62d3761e,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,0.857142857142857,0.9,exact,canonical-source,profile_2cd29242487b8e61b4d8196f,off,off,false,none,,false,2026-09-02T21:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,swe-bench-verified,SWE-bench Verified,coding-agent,412c088b236701cd25bd8bc7207cf2a2e23c1c8e4cb7bb853765852ab755ed8f,8503e17266e1a189664e45a9176b59aa1f6a69c48be8ae8bdcd540e940ba53f8,26fefc2a52f913207707506ebc6c03be814a8c79092be52a1fc37512b171e006,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,70.0,70.0,exact,canonical-source,profile_7c98370c510777de07e3c586,reported,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,swe-rebench-v2,SWE-rebench v2,coding-agent,9128225220206b5af7f3d323af5efc1bc84da52fac6411591d0294df0598b8a7,a8ee5648c3d3e79f9d95cb5c90f39cbe20fa61ab8fa3ff19a3095c5652a3b62c,bb1f0b8a3d0323016acc700587394c23b4e81d4ee438cfee2e8c57d43f6756c2,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,31.171171171171174,31.17,exact,canonical-source,profile_884552ea54f553fb70208863,reported,unknown,,unknown,,false,2026-07-01T00:00:00Z,fresh,123,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,deepswe,DeepSWE,coding-agent,225f38242c6a8b98e93a3fd3eb006df87e5ee5cc491d361e8f189cf6e9171338,5f2dbcd294170afb51f55ec88b0f7fadbc44535eb001ded628a4aa487fba635e,a6e881fb08d7f02373b7ec6ca1562ac9c2dd115460a20ee27445ba6f54a80701,release_7334bb741d8de260846d3128,DeepSWE 1.1 · Qwen3.8 comparison run,qwen-vendor-comparison,resolved_rate,percent,true,13.3,13.3,other-supplemental,supplemental-source,profile_a26de80da0b74a12cbb6375e,reported,unknown,,unknown,,false,2026-08-14,fresh,79,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-8-27b-card,Qwen3.8-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.8-27B
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,livecodebench,LiveCodeBench,coding,64c948e386bdb00ddb89d83bd46ac3dc25dd49bc6006afc7d0be38ddaa778605,c76090d044bead513df1b4f32d08f0b5ad027768854fa30d3d5a649e6ff6f80a,76fab4a2249a73a0e1e012d538500edf88277ec3f4702daebc6bbf5614eec7c3,release_b530ca4da413885c54f52baa,v6,qwen-vendor-reported,pass@1,percent,true,83.9,83.9,other-supplemental,supplemental-source,profile_609752b0f7bd8a6341aa27a9,reported,unknown,,unknown,,false,2026-04-22,stale,193,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-27b-card,Qwen3.6-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-27B
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,scicode,SciCode,coding,0df13a604e53a8430d370de70d7cfa8cdc7024fb5f99e669b5272a26962d5d9f,9c35b67d7aad2943bf5923f74ec0dd7ec0731b3daa073f64d54daf2e87a9abe6,8f32a451a8057bab01b362b93fb8208377ff04c6702857bd22fd3c2677f8247c,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,37.2685185185185,37.3,exact,canonical-source,profile_2cd29242487b8e61b4d8196f,off,off,false,none,,false,2026-09-02T21:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,vibe-code-bench,Vibe Code Bench,coding-agent,36ff10ff8243e1102322dc2aa8602f245657fe7c12330d9923cf4ab99b52dcd7,95fe34c3bc435946ec4317a0449386f59448f16cd7192a7ed7b589edbbf71304,51c6603f5bb1fd0d8c9f1b3ab0d5024e4bbc0896e59cba3c2e578cdd46b57b3f,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,11.941,11.9,exact,canonical-source,profile_0292f7dcc0ed61ad2140555d,reported,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,4bc646dffb022419cf28fe842bc2b23764389437a1ae4e90464e5eb52be7996a,945a0cdd4f86b0ec7e88d0bec3eba2736717861def6b10f9ead39f1f3239d776,662a44299f94415dfc1e341230d74ca6dcabbd11203965171268395946a14942,release_c53c830678318dc1763793c8,Terminal-Bench 2.1 (Artificial Analysis),terminus-2-89-tasks,pass_at_1,percent,true,60.674157,60.7,other-supplemental,supplemental-source,profile_532636b8035b61ccdae69abd,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,3e7649a246d41dd5ec45330828a2a485f0bb773f3e37dceb678061f2d564d1f1,b843dd6ef33f576977d995a9983c8c40a6470c4c3afaac0a90f6a3358ccefcfc,498e8600e17e069a73abb7d078bcd64e4ff805ba73614f86234dd512307956bf,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,16.701031,16.7,exact,canonical-source,profile_532636b8035b61ccdae69abd,adaptive,adaptive,true,adaptive,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,skillsbench,SkillsBench v1.1,tool-use,596c96a04adbcb7e836644982f0cf40987a22d7b317f5814b1a7c369d1847786,9aded900bc59f172374df6f9adce1e88e1babc87c4cd2e4cce3ca83eec8a1290,0075a32b62b602932d1572b58a2560fe97ac92d58166e3cd33e67ab077a6bac5,release_5af9fe5a9d5e275323798c8b,SkillsBench · Qwen 78-task self-contained subset,self-contained-78-no-api-subset,pass_rate,percent,true,48.2,48.2,other-supplemental,supplemental-source,profile_609752b0f7bd8a6341aa27a9,reported,unknown,,unknown,,false,2026-04-22,stale,193,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-27b-card,Qwen3.6-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-27B
alibaba/qwen3-6-27b,Qwen 3.6 27B,Alibaba,open,82,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,a776ebcac947dde1a75991f538dd26602bf6ca8a79ae0115321d8dac6442770d,0848e8979c386e51d4d5df0a4d9e447830def42eb0af805d33d6c0963d722bb2,e3838f4a1e988f8f02ed6b6bfb9a1c6501bc2659ffab211ee4f3db2c168ef387,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,992.98,993,exact,canonical-source,profile_532636b8035b61ccdae69abd,adaptive,adaptive,true,adaptive,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,epoch-eci,Epoch Capabilities Index,capability-index,46b81fa5253f0e9515370cfb87f1b9cf4abfe2c13e4c3105ac34663273c7893c,5ce4537cbdcd3f3c545ede72256232504098d726bbacd50b90e91b335fd5fc68,a89792d48b5dc942994340ac6ae5a601300922e7517257a79de090e954aea31b,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,143.81,143.81,exact,canonical-source,profile_ec15963478c8f005890aa25e,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,b21193e81d616300064c35ece117e7f54bf8d7447c0b4aaa4c2a741fd021cbce,05d26df8485e1bb6d2a57967a93f419a9c8f9261fc1d688f8b5723161d31eabb,e1f956386a97e6ba42a3c396757902d19f19cdabb10fcc91e345af5c771a2890,release_fa9e908ac04b067eb00182d4,HLE official leaderboard,published-leaderboard,accuracy,percent,true,21.4,21.4,other-supplemental,supplemental-source,profile_e4a5a6dee408943601c0a8aa,reported,unknown,,unknown,,false,2026-08-29T23:52:43Z,fresh,63,1,A,1,A,5,C,self-reported,6,self-reported,hle,Humanity's Last Exam,https://huggingface.co/api/datasets/cais/hle/leaderboard?limit=200
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,7cc2ae8b7459db7b59b32c83b22c207433b056f9c2768232b3ae5a3cd4374639,dbabf752d251a308cece4269d7eef5939557386beaa126b75a7006a076a9d7af,7d249a0bc527bb3d21e418292d87d73bd092ea0ac3cbb03fc81ef2633cd19ac4,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,22.242817,22.2,exact,canonical-source,profile_fa62330ab4b31843f371c58b,adaptive,adaptive,true,adaptive,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,gpqa-diamond,GPQA Diamond,science,25f45c87c06458d5f804bb2501f1b2191572d67761528da97aa3e238f6d78866,337025d3ca924fa01952a4dc5f3b3feb3c95d29927175739c5d941d3ab57b99e,43eaf88ff9d29470bcda8b94c866270080c84928f10fd2c2061690b79a22d7c3,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,84.141414,84.1,other-supplemental,supplemental-source,profile_fa62330ab4b31843f371c58b,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,mmlu-pro,MMLU-Pro,knowledge,56025d5f422d3e5038eef6ead2713c289c780a06680a3d3945ecc5b5b549fd3b,b3f100c62868a333f56d5066df5805ed52221f39edcd19b44a6eab2ef379e226,70c7bbff02e177f65267ba3c33be7b6f38096fbaac03c9cc970c7210519303fa,release_95dc04609d027072fa229d3b,MMLU-Pro · Qwen vendor run,vendor-card-protocol-undisclosed,accuracy,percent,true,85.2,85.2,other-supplemental,supplemental-source,profile_771e537906e49bf2223bf52f,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,1af8bbdd697a14e1d2e7541b9b95099768aacc4443330974b919a103914ff6b8,318f534cb9713c7d6cb547575e8ba8d4be46c4cc1b94215f470a03ab237ce952,8b33069dabc4dc7d46ccd6d6ba796cc168c2e534203b12e04d74fff3b0ea4446,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,86.66666666666667,86.7,exact,canonical-source,profile_20fa5f508edb624d3a1dc2c8,reported,unknown,,unknown,,false,2026-08-07T00:03:16.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,frontiermath,FrontierMath T1–3,mathematics,3b993b76c145599ce6dfb3e7bb2e0e2643fa18bc3c9f0b1aa7f5251dfcb4725d,71bfa7e84d6757644a0ad5c45fb058d425a5a3e2c9fa4bc7775374f07411bec8,b4690aeb54cab157351a24640565f368ae0d16229c6fab0cf18e044e124a89f5,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,20.350877192982455,20.4,exact,canonical-source,profile_c69dc5716abeec6f1e0c8be8,off,off,false,none,,false,2026-08-30T22:34:37.000Z,fresh,62,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,mmmu-pro,MMMU-Pro,multimodal,9e5c24afe2e40d7efedbc391e9a7976a4f8573e4b54852b749e1857967d75817,f7a48aaa1c89cc1723ddcb1735b6aa5ddbbb9108cbed7664030e63632c2ece7a,409e11545fbffec5513b54c0e3b8bfa2516be53601cb2eafbaebfb23227c8db5,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,75.028902,75.0,other-supplemental,supplemental-source,profile_fa62330ab4b31843f371c58b,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,critpt,CritPt,science,dec612a45aed84b6577e8b7d5afb09899942d9351876d41b2ce4f6242d4c53a0,d99a4738298837abbcc20ae816e0793a5ca04ecdc7aece66701406bc9bd6fba2,6de27bf2f34049fa14234cab6a3d74f4017d0be8a98cfeb452d9261292b3e7b9,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,0.0,0.0,exact,canonical-source,profile_c69dc5716abeec6f1e0c8be8,off,off,false,none,,false,2026-09-02T21:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,swe-bench-verified,SWE-bench Verified,coding-agent,224c0cb53943005091adb7877073c5fe77eeb15b2fe1e90b5ca844041e515ec4,a4416af63b6b3d09d84ba5e68c68af087f3b48f44b4fccf015387f058218bb32,dad3fccd5f7c7052561a2997a748ad0d5aa39550e8421c973d8656bf054626e7,release_9162beb27813d52a46a20a9f,SWE-bench Verified · Qwen vendor protocol,qwen-internal-agent-run,resolved_rate,percent,true,73.4,73.4,other-supplemental,supplemental-source,profile_771e537906e49bf2223bf52f,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,swe-rebench-v2,SWE-rebench v2,coding-agent,277f0c7209140553e497ec74810fd8734d884598d12ea6206e87ff3a4c868e76,bcc8e41dde7f4b66d8ec829c320dd9761819a8a5c98ef9d10a7318c33383acf4,a399d27e825618cb353e8e50bf75dcd706952ca7575dce1096ed9443bf37b7e7,release_b25eae1fad6b7076bdf70d48,rolling-v2,2026-05-15/2026-07-01,resolved_rate,percent,true,24.684684684684683,24.68,exact,canonical-source,profile_f377cfdc5fe45182122cd035,reported,unknown,,unknown,,false,2026-07-01T00:00:00Z,fresh,123,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,livecodebench,LiveCodeBench,coding,155dae3ff9ef5db154bec3fe3edff55835c072f94dbe16df72e696ffc9ff9ad4,a3203eca6c4217aa2021c6b546a8d783f3d9fb86d5e0fa8725630b7ccfaf72c8,c5897a9b1049c59b6cd6b78df57d0085d6d80e4198b3402492e385b10f29623f,release_b530ca4da413885c54f52baa,v6,qwen-vendor-reported,pass@1,percent,true,80.4,80.4,other-supplemental,supplemental-source,profile_771e537906e49bf2223bf52f,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,scicode,SciCode,coding,dc92775b61568a35938a8683e22011f60d19b3f40a3cfdcdc69e4b9d8290cfcd,12083a6beec638d308dee12ffd2e61b4d27a6aa48fc0a9d37b9e8dc0b0b4cbc7,766e63482093f225aa48b8c9dcd6ae1401800e990b7a82beb625e35e35b8bbf7,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,36.5740740740741,36.6,exact,canonical-source,profile_9f8b471c8142f59bb8b69a15,reported,unknown,,unknown,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,webdev-arena,Text Arena · Coding,coding,8f5fd428cec0612b9e31b687887691cf101d7491b75de81e5fd057da9cea81f2,0a6740bfe0003766d05a2be59d494b06fa9d5ed3ce16fecfec6c6da4f6fc00a5,c9662d1a8fa00d0b928c4439fae85fbe4d0b0b5bf5f3915e743eece8b290e8f5,release_65fc140e7ffbb7bf62021f84,Text Arena Coding current pool,text-arena-coding-bradley-terry,bradley_terry_rating,points,true,1250.15,1250.15,exact,canonical-source,profile_7831c3668b6ad545a4b181f8,reported,unknown,,unknown,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-webdev-arena,Epoch AI mirror · Text Arena (Coding),https://epoch.ai/benchmarks/webdev-arena
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,a7c1f69834b429ed2617399373cecabff5182d6bf2a5e50cb15b7c1fb44ec01a,809e4b563e243904bafd67a0e2403e35e7adcf5f2ff60a91de5fc994f9f52f1f,a36d5f73d6e5a36197320d6f5a15cecaf242e49008583c6dc40589b9d1e92211,release_c53c830678318dc1763793c8,Terminal-Bench 2.1 (Artificial Analysis),terminus-2-89-tasks,pass_at_1,percent,true,44.94382,44.9,other-supplemental,supplemental-source,profile_fa62330ab4b31843f371c58b,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,0dde2bcdeb6bd4e6db7304498ec0905a98de2b97431078c3e4d134e1e8d15bfc,c006572d4493d234c814b6404dcbb94ae99299a2de00c6b52af30946e625cf7f,14947b626dca43242a89569994807d9d005f2ff17b42177d4963c9ed18fd3609,release_b7d9ab744ffac0cb856e114d,Terminal-Bench 4.0 (Artificial Analysis),mini-swe-agent-66-tasks-500-steps,pass_at_1,percent,true,0.0,0.0,other-supplemental,supplemental-source,profile_fa62330ab4b31843f371c58b,adaptive,adaptive,true,adaptive,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,9ba30ba3e9d8a0cba7896284c6b52afa4f8e0b77714096b816931af4eecc5a6b,0d5b68e31981ac4fedd61c3e8491a5c3c9a47200b159582942a6cfc9ec3ba275,fe6ecc7d35a7d6b511e8c0e144cbe3f1174e02bfef493c3afd91004448589cdc,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,9.278351,9.3,exact,canonical-source,profile_fa62330ab4b31843f371c58b,adaptive,adaptive,true,adaptive,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,skillsbench,SkillsBench v1.1,tool-use,7c4f2d563f9e4f620d204582b31e1e91f63ef7c14f7cefb4c42bc5275d6c28eb,ac46754bd3952d481779be1a27c2f62f553c72965f6be0ea0f7234bf91c87341,1ef4df8bbd26617f6a97c4748bab4ad0b89cb62c839ee783c7a5b25a8ac69428,release_5af9fe5a9d5e275323798c8b,SkillsBench · Qwen 78-task self-contained subset,self-contained-78-no-api-subset,pass_rate,percent,true,28.7,28.7,other-supplemental,supplemental-source,profile_771e537906e49bf2223bf52f,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,4481172c8ecab8af86b10b2544514a2e5b0f8d9ab7683e51a5500dbe8f19759c,c527342f4cf715a1fac48c75dcc24534b6f8a6c29199ce2ed185df2fa923a9e8,7c52cbe81c74dcf38f40cee2c97ab50726fb2468ae1e1df76082a1c916930c1a,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,899.46,899,exact,canonical-source,profile_fa62330ab4b31843f371c58b,adaptive,adaptive,true,adaptive,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,e574e87e5070501a934849b975a8bcea9b21372b779ddc0338d4274b5ad49904,0eb52403dc1e2c99a89458965f90edefc52618f10f346ce6633a4e5bbf742904,ac033ae84d9e8c3792065e58a44a15f3d80b846134d957b52b2b43ded2458ebb,release_c6e5abafbb7bffc0975932de,TAU3-Bench · Qwen vendor run,qwen-official-user-model-bm25,pass_rate,percent,true,67.2,67.2,other-supplemental,supplemental-source,profile_771e537906e49bf2223bf52f,reported,unknown,,unknown,,false,2026-04-16,stale,199,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-35b-a3b-card,Qwen3.6-35B-A3B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-35B-A3B
alibaba/qwen3-6-35b-a3b,Qwen 3.6 35B-A3B,Alibaba,open,83,weirdml,WeirdML,coding,6d19c9fc3377d2468854ecbfb8aba4529637ccb9f113c1a1875f08dc406d6a70,d3ba5a61e6a9e3978f73df75ca41761992a18a09b6557ffaf88d9fa8c07ac2c4,c6c795b7e9a1dbcd297aa30fcc438a0b5d01f63aeb58081a4b8c2776603b447f,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,34.49,34.5,exact,canonical-source,profile_766e368be42d8b18446b5ede,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,livebench,LiveBench,general,23c8f637313e13ee86c5a006d08a712687ca350daea18b06e95caae320ef1ac4,fc84ad62421809cc0539b9568c74d402790d4cebc56de197eeaee90db83719a9,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,76.19415476190476,76.19,exact,canonical-source,profile_d95c76a014239d533b5e8b3f,reported,unknown,,unknown,,false,2026-08-27T22:30:36Z,fresh,65,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,49d0aee1d3d042d3f65fe4f1ee5742ec7a78f5123e770f6928e9284a4cc7696e,6d626bd77f15e5242b6d18759f61c6bfbf2db0fbd66caf747183fcf5f773065f,70e596e918ba6dfcb89e204cfc4f567f7cfee53e41c3d48d086272421582cf13,release_287aa4bcb31b5b1ba26e562a,HLE · Qwen vendor run,vendor-card-protocol-undisclosed,accuracy,percent,true,35.9,35.9,other-supplemental,supplemental-source,profile_0ee939ebaa5890202c8663ac,reported,unknown,,unknown,,false,2026-08-27,fresh,66,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-8-flash-card,Qwen3.8-Flash-Next official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.8-Flash-Next
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,e66906c8a3ec83ecf82f6cf8b3a0179d857965a793066f094eb5e2075a4869a6,f3abad538a865df0d32eda6ecf280877592f3478b41513a3c3a989870fa6f130,fa3a16087856d6b370a7cc0bddb3305dd311110562c9d2678fa42f46fec06978,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,38.044486,38.0,exact,canonical-source,profile_a9a476e7573bba3097d22172,reported,unknown,,unknown,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,gpqa-diamond,GPQA Diamond,science,be4db36753516bfad62409acb8c07fbf32cfed02c9fd032f03667befd9d6414c,13c5d95f1c9407120e26f3dc87410ea852f95a2f522e51ca8454802300812007,96a8d2d312a7017af06c07868a02badf24feae0b6af2f80ef79cfd4a09f2bcbe,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,92.323232,92.3,other-supplemental,supplemental-source,profile_a9a476e7573bba3097d22172,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,mmmu-pro,MMMU-Pro,multimodal,8831df20f5f870224c6c67ade64ea474cafb4a55ce8cbb74fac2ac1e601f674e,043117908cf30ce05b0bf2d02b0ecb53964675b543471b0e39000be5f30fd087,343457ef16609f777d8a29321f25ebd88d24f7451aa6ce6625c800d64e623b43,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,79.768786,79.8,other-supplemental,supplemental-source,profile_a9a476e7573bba3097d22172,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,critpt,CritPt,science,c1c7a90be90ec6878521ad2562c12ac18a3e6193da387e3288008fe971fa530b,4e4dce762d957c28dfd118dd80d68c64f5236000eb9b053b219f419e15e2dbeb,95e3251a48840fe43ec7d3a35c85d33000956d934cf5658c049514852a511dc7,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,11.142857,11.1,exact,canonical-source,profile_a9a476e7573bba3097d22172,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,deepswe,DeepSWE,coding-agent,e90a3198f6d981321746ff06de256d780783c636f766d15c1c8338fa409b1cdb,949a28e6130f4052403eb43c6cdcdf4c1b49a73e88bd5103937d5640882024fc,8952100b315bc05cb6845e1f3c7939f39264ad2d12eab342c072a06b1c4f1351,release_12e8a62d68d933c29dcdf020,DeepSWE 1.1 vendor run,vendor-model-card-run,resolved_rate,percent,true,58.7,58.7,other-supplemental,supplemental-source,profile_0ee939ebaa5890202c8663ac,reported,unknown,,unknown,,false,2026-08-27,fresh,66,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-8-flash-card,Qwen3.8-Flash-Next official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.8-Flash-Next
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,scicode,SciCode,coding,06ac9d6efd71ee7dc92d0a31174fb41bdaa7d310d16096e1422b300d7507efd0,bd88a3761ecdf0891f381973cd8a40d75525a66f663eb7f802057852718dc180,e8b485c58fee3e97f4fe7a224620a310102d3c6952e5799ecc39671660311646,release_de40c66bd40a7b5483b40f9a,SciCode v1.0.1 (Artificial Analysis),scientific-research-coding-subproblems,subproblem_accuracy,percent,true,50.578704,50.6,other-supplemental,supplemental-source,profile_a9a476e7573bba3097d22172,reported,unknown,,unknown,,false,2026-09-05T17:21:27Z,fresh,56,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,webdev-arena,Text Arena · Coding,coding,f65221593ad2b841b953212b5455dde404372e9545456c5d91857a89878ae0a9,f6c917cae705f3c8294efea0635e591eef12801d3d3ceabd2f695001186a5080,b8edd355d20d33258f3eca81e0d0b87541d0515330d92d6324b937db2ddf48a2,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1628.3220907822047,1628,exact,canonical-source,profile_c9c41d02d7ef3f31e9d42eeb,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,27aee8d88572b62f0a90e0f81d286ba3499ffd2d0f1189dd2863f178a77443cd,e7be9ef0ed2a31067b66233bb66caa3b509e4aab835e956922b3479d32b4cf8b,5f6fb3bea08984c87ec14fba7756a8d20232a7ac9b38f3bcc2ec256472771f93,release_c53c830678318dc1763793c8,Terminal-Bench 2.1 (Artificial Analysis),terminus-2-89-tasks,pass_at_1,percent,true,86.142322,86.1,other-supplemental,supplemental-source,profile_a9a476e7573bba3097d22172,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,09dabf754c08ced5d1edfd1da8b8beed059c2ea2c064f6cbb924df7058fa3818,f30f1bb94f7c4b6bf8ca555c176334248fbc9132c01e56d3f43c8ea0d9bdcca5,074587c74ddeba8d16bc4aa58ec9ec9ee03b64ba6db2db5bfab96c64f3fd53ef,release_b7d9ab744ffac0cb856e114d,Terminal-Bench 4.0 (Artificial Analysis),mini-swe-agent-66-tasks-500-steps,pass_at_1,percent,true,25.252525,25.3,other-supplemental,supplemental-source,profile_a9a476e7573bba3097d22172,reported,unknown,,unknown,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,f1d9597544051938bf47a0a4c142e5e9f3d777548c71c01d56fc5e5e1d6b1173,38e33aa0e5eb214e27961806fbbd0bebdb5e6c7dba603dbdbff29e7cbe432581,fdc0aeb253baf28cd463d39c9ebab3b9206e0bdaa677e65a59a9743cd272bc31,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,45.360825,45.4,exact,canonical-source,profile_a9a476e7573bba3097d22172,reported,unknown,,unknown,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,arena-agent,LMArena Agent Arena,multi-turn-agent,442afa4ba4c3760951bcaac67f0947e491ad2ea762e3bf29fac3da7ea9e8c811,2f1a43e4ebfcbef878f6d530f831a8b290759db10ca2647b3a47dbd96911a3f6,1ba5ed3ee1b56f6117b4921d36f1d7ec9f63734e33fcc1eff4067e87fb591576,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.04022661938759629,'-0.0402,exact,source-specific,profile_905688dc095b693c8f3fab53,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-8-flash-next,Qwen 3.8 Flash Next,Alibaba,open,84,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,fbdb49cc508f7b40ca48b89f96eb7c310966cb0dea77b9fa1c530ad0c563ec53,a03bf184cc44b71054168f8885bcce77784aa43ef879a2e170aa5b4529afab18,fb4990f26a0e638d811ef69156d86fa985f98d759defc5e8df18fad20f1d90d1,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1635.78,1636,exact,canonical-source,profile_a9a476e7573bba3097d22172,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,epoch-eci,Epoch Capabilities Index,capability-index,1ceed7e7b51819379dc1588811d41ccbc9dd15b11331b12250492317bfd8b202,697ca445bd3dbc126e32c8fb577ff07061db833c8eeced0c788764495f1ad9a0,3f8096875b85336944d40d3b06a9b502e3c6d15ea3eef0543a5fb399c882b1d5,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,143.08,143.08,exact,canonical-source,profile_5df12bd86522bd0fba242adc,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,33b6281ff3e4ea642abf9498634127dc6495925dfb92c55ccb9411fd7123b2a7,f29f4178978b7af215f39660737eae7e7c3029f1ce53d84b4ac3f50fbb003879,9bf33a886a0f23c36ca21f71357edb46598a0aa53c037d5b817890035204e13e,release_9b8cd6ba0a835eb2da22bb93,HLE,google-vendor-reported,accuracy,percent,true,19.5,19.5,other-supplemental,supplemental-source,profile_e32b67ae6bf59f0434d4b67f,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-31b-card,Gemma 4 31B IT official model card,https://huggingface.co/google/gemma-4-31B-it
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,simpleqa-verified,SimpleQA Verified,factuality,46b98fe9db692e65b4555920d3b384862d63ec16a682254b9ad8bd2cd085fa31,3b8e30cc051012cca3f62f02e0a30dddd7c20ac090da11df88f85f53049de715,3ac8135ce62560dc2e1a768e0d30d97275fa17858342f1639436428e2c14b10c,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,10.4,10.4,exact,canonical-source,profile_3bc4445984e2cfe1b5212307,reported,unknown,,unknown,,false,2026-08-27T19:30:07.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,gpqa-diamond,GPQA Diamond,science,d376f4e49689556b76201a01d25993ffb542a4d059898a61546c91a38589a4a1,6e02d42e2d21ac45fc5435e3ce78467997da817c01fb0836e6604e00ae06fcc5,32b3f877f64442a1212d93dcbfb81f5c80369c109f440911702295426a25237a,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,75.75757575757575,75.8,other-supplemental,supplemental-source,profile_af2ff99bab99820d4a9eb678,minimal,minimal,true,fixed_effort,,false,2026-08-06T23:59:38.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,mmlu-pro,MMLU-Pro,knowledge,4f9949cda683889721cbde50c847a62adada4351046b79163a263f6f7c97fe14,47a3f272644a74934efee821a31ac6b49d38ab79f6db5c4b81a7d28d95df49e4,78530b6152fd4eaecf30afb5b591e4d005694e92002f866fb1149eea1731e217,release_cbef64e634e73d9217bfe565,MMLU-Pro,google-vendor-reported,accuracy,percent,true,85.2,85.2,other-supplemental,supplemental-source,profile_e32b67ae6bf59f0434d4b67f,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-31b-card,Gemma 4 31B IT official model card,https://huggingface.co/google/gemma-4-31B-it
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,0605537700019c3c627460b1cf538754616d4b111b405107e658e76cd26c2308,06fe7f3aa09c562ad2c99d740b3c4fb874387ae04c2da7d0e67b9a2810e49a24,07e5cbaa45798df58d2463abaff06d5b328765e707e6e30a522c548f450919c7,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,73.33333333333333,73.3,exact,canonical-source,profile_af2ff99bab99820d4a9eb678,minimal,minimal,true,fixed_effort,,false,2026-08-06T23:59:45.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,mmmu-pro,MMMU-Pro,multimodal,b79e7440839dbcf84a17ad59d2677f7937dc77a425fcc4325cca5892dd93c776,6f348d5593d6801efe9aeea7e579c7a2ddb03d14b3b9a7f12e2f1b68535d0b4d,b4ea8f2c161f678a1cd9587ed3df25a8dd6a87cbc1e1c751bec680b93db62fb0,release_6f3fd0029277bfcad78a10b3,MMMU-Pro,google-vendor-reported,accuracy,percent,true,76.9,76.9,other-supplemental,supplemental-source,profile_e32b67ae6bf59f0434d4b67f,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-31b-card,Gemma 4 31B IT official model card,https://huggingface.co/google/gemma-4-31B-it
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,critpt,CritPt,science,b9dda1a7a1bcfa245a0b27dc5b5d5a94adfb053338ef6ad5e0448149b8de200b,5abb5f89b01ff96359cd1a5b1017da7bcb7971173ab2e0595031693a7205d5b7,0cfe56e5bc37e596794d50564f26128c19f9853fed0f910971fbed831d41d643,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,1.42857142857143,1.4,exact,canonical-source,profile_7d82a165512b034548c396d2,reported,unknown,,unknown,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,ifbench,IFBench,instruction-following,fd2bcf07c727dcc8176a3dcd5cf4a9114db31356efb3bec74b4daed4945ffc16,6bc1a87a33305bba5e345913c69faa169b17aca9640083ab6484f16524d31678,6cfb052f9166158b3e6c3af96acd982cbea69c72aac5685f77a8a96d272bc4cb,release_4adf348ac2f76539ed7ccb10,swallow-evaluation-instruct,ifbench-test-300-prompts-344-instructions,inst_level_strict_acc,percent,true,76.5,76.5,exact,canonical-source,profile_82d7b860b7defe3c83636cc3,adaptive,adaptive,true,adaptive,,false,2026-09-30T18:53:19Z,fresh,31,2,B,2,B,0,A+,official-leaderboard,6,official-leaderboard,swallow-llm-leaderboard,Swallow LLM Leaderboard,https://raw.githubusercontent.com/swallow-llm/leaderboard/c58a13a852eba061d639af7caf4ddbdff33b36e3/swallow_leaderboard_post_en.js
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,livecodebench,LiveCodeBench,coding,2f34f468b18cbd0c11832768d6dd2e9ea628350f6de035ca9b3dfe45f767d6da,eaf6d59bbcdd16bcaac3a5ac06f41977df3acc8aeadd3ee2df9e26b99c1817a6,6d83bbcf136942f9b5279b1d5eef6f61d1c43f1bfd5d2d2a9b7aee53f4a643f4,release_1d1cf16bb310d2fb0576bb02,v6,google-vendor-reported,pass@1,percent,true,80.0,80.0,other-supplemental,supplemental-source,profile_e32b67ae6bf59f0434d4b67f,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-31b-card,Gemma 4 31B IT official model card,https://huggingface.co/google/gemma-4-31B-it
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,scicode,SciCode,coding,29e576a435b30fd9c90afaebf4e3efb898912a3c6efc730c34909186f0ad2b5f,7d20e6093f81c0b03b0b41d6f6b7bf2de73b47c7bcc6473baf84595b2387dee3,78047ad15e5eb077962e61f38f7459e1eec7493ed9d02e1063369fc7aa5a19be,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,45.4861111111111,45.5,exact,canonical-source,profile_e139c296e22aa94481af21b8,reported,unknown,,unknown,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,ale-bench,ALE-Bench,coding,39dd03db89da3440a278ad54919e1d79954ae4219040669ee001e4c036a21996,c756e2d8be60aba061b67a1a95cfa26df11fcd01919b89378affb878ea1959db,7ad5ec9cfe49e1d8ab07f2b42696a1b26507dbdcb90832dad73af780b7a383e0,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,925.5,925.50,exact,canonical-source,profile_9bb470d85c107daac9f268e7,reported,unknown,,unknown,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,webdev-arena,Text Arena · Coding,coding,b611de399190d651846614a7f08c3e9fba3184a2fe2a71bb514d1b2e9abc929b,196e1911cce5af03af1085f765d196a61c96c3a4036df0dba4f332aacec7cc2f,a1493a09d9576bf0250a9a2dc1b687cc35f4f06bd586fe3e0007f0ff135aad04,release_65fc140e7ffbb7bf62021f84,Text Arena Coding current pool,text-arena-coding-bradley-terry,bradley_terry_rating,points,true,1364.13,1364.13,exact,canonical-source,profile_d9228fbf3d63a4e480591aec,reported,unknown,,unknown,,false,2026-09-30T18:53:15Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-webdev-arena,Epoch AI mirror · Text Arena (Coding),https://epoch.ai/benchmarks/webdev-arena
google/gemma-4-31b-it,Gemma 4 31B IT,Google,open,85,weirdml,WeirdML,coding,70a3a8f413de5dac2933b048e958c39dbb1e9ae1f2278a9e40f40e96e8db1de4,a1a64ae228d171aba40f706420edf01a6d9e2455f8953c138323916d1b6c504b,18ea39e9a2f389d9640f329d7b2fc41406f5d7e4329aa71c69b1482d9dbd6ece,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,52.26,52.3,exact,canonical-source,profile_e9bf89560c783f7a74d891aa,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,epoch-eci,Epoch Capabilities Index,capability-index,e6c93b73d48c4aa2c1764c0880e9b031be3349355127f7e25c1b14d37be0e4f5,9d68134a95eea497d13fac987fedfaa752a19e6c0271d81676b162e913bb9e1e,a7bbe8483e8615ef42616e54afc59e20ef74d773969cb6f3b6143ecabe42cbf0,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,139.37,139.37,exact,canonical-source,profile_92d32148ceeb2d09e13b1a18,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,b7868f0886e8d8083adca18701114284576569c024b57263585dc4cd826d0746,4d5d33513341af3eed2237e567a0e7fefc4e2bc053508272f1525ea6a37d5314,17211bf56287f72c07f53c3763c7956d1c2da66d284e1722f2a454e56e656c4e,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,14.921223,14.9,exact,canonical-source,profile_128d9317b7e94f2fdf3d00ed,adaptive,adaptive,true,adaptive,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,gpqa-diamond,GPQA Diamond,science,0187438286baa92acaa9bf9bfa0eb73ad8d08e7958d3bdbe99f153dab485911f,5f0c55b62d650d789b21fb0fa27e3562d90a195966664b769afd07a0e50a09a4,4232061a938da4e3519e29afe67de368f790079a5b3f48f6520b7daac397041d,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,80.606061,80.6,other-supplemental,supplemental-source,profile_128d9317b7e94f2fdf3d00ed,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,mmlu-pro,MMLU-Pro,knowledge,9e3c9648fd59b167e1c8c0819cebc876bd729996c0c8cd50f46b075e45062476,8eb7cf6080d4d5949b44a517ebc2102339590daeed7cf8a362deae972c072c4b,67b82b3d54800e8378e5357b208e7d192247e30225c7df8f7d7fbaa13b033818,release_cd90aeaaf8e38ae0a7eb7edd,MMLU-Pro,qwen-vendor-reported,accuracy,percent,true,82.5,82.5,other-supplemental,supplemental-source,profile_b87cc44f5a3902b1b71e87ed,reported,unknown,,unknown,,false,2026-03-02,stale,244,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-9b-card,Qwen3.5-9B official model card,https://huggingface.co/Qwen/Qwen3.5-9B
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,ca8befb728e3c92ea525932d20b9a10711cf3fa0e8958bfd6bc54859cf7a416f,7dc6abd765e7f88864fe6b49795d01ca3246f228c7e82768d1851cf1b65d2f0c,05e2ba21e785d3047e3d2cac876f614eca4581d545cf95a07baa82d5f5802307,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,61.66666666666667,61.7,exact,canonical-source,profile_4779204b36659935aea6f3dc,off,off,false,none,,false,2026-08-27T14:31:08.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,mmmu-pro,MMMU-Pro,multimodal,446c47973eb0727194a12afe4f071d86c294c8412e248852d408a5cd64b748ff,2f45ab3cdfed1b31d8a3b6691304a87de5ab071492535627a758c1bba67407bf,34dd4fab6dce26b8d061080a94c98fe213c3d3ab32dd98def5e7ff099cbd78bd,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,69.248555,69.2,other-supplemental,supplemental-source,profile_128d9317b7e94f2fdf3d00ed,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,critpt,CritPt,science,cea0225f303bedf6b2830c4bc9cb6c017aed351e783044cac3f076f0d2a91b99,e461ca18d9ebecf8f2718238ce01ad4853e92942f8991720a1fdc2594584f590,3963f6d4480ad557ed35cc4768ce48d66f48efb32f20ceb414f149f8b9ab5426,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,0.571428571428571,0.6,exact,canonical-source,profile_4779204b36659935aea6f3dc,off,off,false,none,,false,2026-10-28T21:25:03Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,ifbench,IFBench,instruction-following,e22c80f41d0986c1a1b38bc1bbd57a06b2db83c2408e5c8d7e6f55dc74ae1b75,d257907fc25eed9b044ee68378705daae998f98425c18d8b0f7f61d2fd2c23b6,73ea367f7a86235a5d384e9dce2985cfcbe634f786e9dcc20c22225705467b05,release_4adf348ac2f76539ed7ccb10,swallow-evaluation-instruct,ifbench-test-300-prompts-344-instructions,inst_level_strict_acc,percent,true,48.3,48.3,exact,canonical-source,profile_128d9317b7e94f2fdf3d00ed,adaptive,adaptive,true,adaptive,,false,2026-09-30T18:53:19Z,fresh,31,2,B,2,B,0,A+,official-leaderboard,6,official-leaderboard,swallow-llm-leaderboard,Swallow LLM Leaderboard,https://raw.githubusercontent.com/swallow-llm/leaderboard/c58a13a852eba061d639af7caf4ddbdff33b36e3/swallow_leaderboard_post_en.js
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,ifeval,IFEval,instruction-following,b843259777bba2aa4b266d44719f3606efdd96a235abb7aecd488390753637c6,0d4d2a96a79a786100ab483068ec76c74fad3067e340e91b85af0a37a8facba3,c2532bd711b5e3c0b76299a052a42563ef11942e70f026510be845076dcb1554,release_eaa2f0262bb0264860cb2c8e,IFEval,qwen-vendor-reported,accuracy,percent,true,91.5,91.5,other-supplemental,source-specific,profile_b87cc44f5a3902b1b71e87ed,reported,unknown,,unknown,,false,2026-03-02,stale,244,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-9b-card,Qwen3.5-9B official model card,https://huggingface.co/Qwen/Qwen3.5-9B
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,livecodebench,LiveCodeBench,coding,ee0ed12422f2fea70a6d103fe70e39d4e43ab28813161dda0c31e04dcbf7a6dc,a1d652cbfb6a18325c2ff5520422d392a72743d8246741fc8cdc51dfb05127fb,c1d1df0ddb34c3aad1342929c2b8d8d955114aae5c9a0d882037f2e4f2898440,release_b530ca4da413885c54f52baa,v6,qwen-vendor-reported,pass@1,percent,true,65.6,65.6,other-supplemental,supplemental-source,profile_b87cc44f5a3902b1b71e87ed,reported,unknown,,unknown,,false,2026-03-02,stale,244,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-9b-card,Qwen3.5-9B official model card,https://huggingface.co/Qwen/Qwen3.5-9B
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,scicode,SciCode,coding,07f28a554f937caf024252099c601e53e6e968d690391469e6a49522c9f24bb4,f6afd2c6ed210a4258728cb0a47198c2782da22209d675585cbb7ab30a351bbf,ce03d7b06d623fd647ca7fa0f501a311374ee36b58b6abddccbfb58c5e2868bb,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,29.5138888888889,29.5,exact,canonical-source,profile_b8beb3f11c16f95fe7bd0c8e,reported,unknown,,unknown,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,d9456099806ed96af5cef3c547a5a0d5717280bafc52f293e4254adcffd3bc59,395eda4cce822acbdc033aacbe9641d682aa65f4d2a757d6a294f4501dacf1c2,7b8ecb4fba82b9e2c7d2e39b08d1f2ffbc09fa7cb79786b323a00bf93933c07e,release_c53c830678318dc1763793c8,Terminal-Bench 2.1 (Artificial Analysis),terminus-2-89-tasks,pass_at_1,percent,true,29.213483,29.2,other-supplemental,supplemental-source,profile_128d9317b7e94f2fdf3d00ed,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,1bf98ca38122403a56a213a1f77e1d6820f4d532ee8fff704fbd2b09e8e5f5ec,2c892a023c585ae0e99f90aa99b99d7d8b9ef0a1192e49ca77b6117bfda92904,a092f43887d1c2633ca620892aef0f93724cc2ed9ef9f780a5b7cbb58a70b572,release_b7d9ab744ffac0cb856e114d,Terminal-Bench 4.0 (Artificial Analysis),mini-swe-agent-66-tasks-500-steps,pass_at_1,percent,true,0.505051,0.5,other-supplemental,supplemental-source,profile_128d9317b7e94f2fdf3d00ed,adaptive,adaptive,true,adaptive,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,a1f33b3ee1231cdc9df821c82e545b6a0f95d77676f380de099fdc4218fa8f8b,47579c450dc6e32c121a8786b0d2c3f378780975b87843e3704e5de1d7923f4b,40e62be72828723b1479c7f8217fb6bb92767f840aa690e6f01aeef5c118c823,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,7.010309,7.0,exact,canonical-source,profile_128d9317b7e94f2fdf3d00ed,adaptive,adaptive,true,adaptive,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,bfcl-v4,Berkeley Function Calling Leaderboard V4,tool-use,486d953a6edb19e2e0e1eac69de127d049a38ad4aa99be8d8dc5659d2a9f9dad,83a334dce84488221936c7a13fadd794bdb17317f106acc1dd477eae4afbd8b5,9d10e802f12470bbac18314ad12a3fbec3074eedd5d2186e2a365133438c65d4,release_367e04d56fad3a6f9849489d,BFCL-V4,qwen-vendor-reported,overall_accuracy,percent,true,66.1,66.1,other-supplemental,supplemental-source,profile_b87cc44f5a3902b1b71e87ed,reported,unknown,,unknown,,false,2026-03-02,stale,244,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-9b-card,Qwen3.5-9B official model card,https://huggingface.co/Qwen/Qwen3.5-9B
alibaba/qwen3-5-9b,Qwen 3.5 9B,Alibaba,open,86,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,cc1b5f17198a94cc14bcd50f89c9574235d58121794c425baeb5990a1fe1f62d,3b9bb4f9da67e2f57c7bc7a75a2c46d7168e31121f6d800134e05512410bebf3,dfc0ccfade597262db190d2a32dc545f2a3978536f008dbe6d9dd0a82d033431,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,426.58,427,exact,canonical-source,profile_128d9317b7e94f2fdf3d00ed,adaptive,adaptive,true,adaptive,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
alibaba/qwen3-14b,Qwen 3 14B,Alibaba,open,87,epoch-eci,Epoch Capabilities Index,capability-index,7bb3fe32ad31a3562cbb1f06e2e058fd4be699b04dc9c5b9893141bab743d13f,d6158b10cf0e12274b1bd78d67a195b492ca7222c342016bf24c82fcdbd3995d,4cadc4bfe54fc11c813b9c46c5584cf989bb2fcebfe4bcec7cd80ff70e752f1b,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,138.2,138.20,exact,canonical-source,profile_de4bc2746c30edb0d1633245,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-14b,Qwen 3 14B,Alibaba,open,87,gpqa-diamond,GPQA Diamond,science,9306dc03b93ef652ec0af01630bb56c0b9cdd6825cd0b1c5f58f22bddea7f9eb,b8901bfcea26f29206029bb1f5287101609e1b44b2899038fa850e6dd0b3ab22,5b280138ae3a1b4b075a4a4d1589cbfc5ce70e41aaa4f0fa1b5d23f94de2d258,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,53.40909090909091,53.4,other-supplemental,supplemental-source,profile_67bb468286cf106525571e45,off,off,false,none,,false,2026-08-28T16:54:57.000Z,fresh,64,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
alibaba/qwen3-14b,Qwen 3 14B,Alibaba,open,87,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,1afdfead017cf75066f1cf69e29c5b53215fec31a39b216a0553eaed681d0f4c,f800b18d1a8b03cf98ff24407319e3e90d7d34838dc18722c0156b660e51532e,82dd0f787fed28a4d341c9c175a630d392f770778c928a6945501cee25d76b47,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,25.833333333333336,25.8,exact,canonical-source,profile_67bb468286cf106525571e45,off,off,false,none,,false,2026-08-30T22:42:57.000Z,fresh,62,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-14b,Qwen 3 14B,Alibaba,open,87,critpt,CritPt,science,50be307c898b725bfa259f8738d0e8eaeb717a93fb7b8ed73b4001c2193c1c03,bbe681072662e313775063854049721b450309f846b1791c245f6966179fe320,9e57af0a6ede2d58f965f4e049058a4a2c8ad42b3a15fd865d21d3d2db170f48,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,0.0,0.0,exact,canonical-source,profile_117d28b5512fedd5597760f4,reported,unknown,,unknown,,false,2026-09-02T21:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
alibaba/qwen3-14b,Qwen 3 14B,Alibaba,open,87,ifbench,IFBench,instruction-following,1c7cdf0ea5925ae7ddcbb9cb7dbbececbbcf8d408fa4522200c1eeb4f2aea355,b34adedafb1fde9d3f452badf0b307a7770fc4c905dcecfaa026dcd714a115b0,6492833b49c17c6db29dd2299cdcda765ec1357b6c689b61d0600cba0e168682,release_4adf348ac2f76539ed7ccb10,swallow-evaluation-instruct,ifbench-test-300-prompts-344-instructions,inst_level_strict_acc,percent,true,32.300000000000004,32.3,exact,canonical-source,profile_c046a0e96997ecfa1f9b15d9,adaptive,adaptive,true,adaptive,,false,2026-09-30T18:53:19Z,fresh,31,2,B,2,B,0,A+,official-leaderboard,6,official-leaderboard,swallow-llm-leaderboard,Swallow LLM Leaderboard,https://raw.githubusercontent.com/swallow-llm/leaderboard/c58a13a852eba061d639af7caf4ddbdff33b36e3/swallow_leaderboard_post_en.js
alibaba/qwen3-14b,Qwen 3 14B,Alibaba,open,87,scicode,SciCode,coding,bf8084b04648bd59fc95510c6db770dc9a5df917095415464e86e648709fc2af,b1fc55f0eb78e35c2afd995737d2799e488745bf30b99af8e07a64a8697c5861,14600173be83f44c993298618540e9f6ba4f2dbd4d19dbdd47f484352f9151ef,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,31.5972222222222,31.6,exact,canonical-source,profile_f72a9c1ff904f7e8fb9475ac,reported,unknown,,unknown,,false,2026-09-02T21:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
alibaba/qwen3-14b,Qwen 3 14B,Alibaba,open,87,bfcl-v4,Berkeley Function Calling Leaderboard V4,tool-use,a4be7722c0edcbeca9ee305400a8f3df1d6a0b321ffa1f9ef51f7ad5c2eac59b,3b39bae4fee1ecfc98e5010cad97c89a5c8d0375f97c6b10e4bf0e29760015dd,6a9c99bccbca11748bdf70a2bdd6eb864726c94bf4a029f84ab95898fe96d192,release_7d2560723dba9d549cd10b25,v4-ede5081a24bc,overall,overall_accuracy,percent,true,41.03,41.03,exact,canonical-source,profile_b2ae49bd61766d7bcecc4aec,reported,unknown,,unknown,,false,2026-08-28T22:03:18Z,fresh,64,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,bfcl-v4,BFCL V4,https://gorilla.cs.berkeley.edu/leaderboard.html
alibaba/qwen3-14b,Qwen 3 14B,Alibaba,open,87,appworld,AppWorld,tool-use,add59be5c620d5f99ea80296255008411c5ed53e69b7eeb132c55657881ad2b5,6a282aa75481b9e6924ce9566869eae06b580de362e97655cf399e9f83b2edb7,3fd2d05d1806fc66922d5d31d1ed44fa2ec28c31ae87e104d2ae6d50e87e1673,release_3fa145932216c0afcdd9dae6,official-c1f56015cf7c,test-challenge-all,task_goal_completion,percent,true,67.6,67.6,exact,source-specific,profile_65e947029c095ac93b048fea,reported,unknown,,unknown,,false,2026-02-15,stale,259,0,A+,0,A+,0,A+,official-leaderboard,0,official-leaderboard,appworld,AppWorld official leaderboard,https://raw.githubusercontent.com/StonyBrookNLP/appworld-leaderboard/c1f56015cf7c3441ff1933f5bfbec879798d7bbe/experiments/outputs/_leaderboard.json
alibaba/qwen3-14b,Qwen 3 14B,Alibaba,open,87,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,8497a24538b3acf91108a6c1cc132fe22891ad95d6bcc1cf4ec85beeb3d29fb4,202091fd9c66be1cb0cb293b3f3c275a104f7c43ade67cef73cad3d3c8d40a6a,99fa32f2d158823ab7e3c18a005a794c66bfa4b940eb6307004a6d2210995c71,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,-17.25,'-17,exact,canonical-source,profile_c046a0e96997ecfa1f9b15d9,adaptive,adaptive,true,adaptive,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
google/gemma-4-12b-it,Gemma 4 12B IT,Google,open,88,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,d18c8a8cd0cca8206b37c89c6780e6765d3e737fdeb12375abd7ee6aa3847428,65a5492c7a96153ccf605341e32e1777d82d4e3af9fe4ab26af91fb02872f5af,c7b0e4330c983c8b61ff308934dd9e46ac24e7b939d101095153586aa5e21a2e,release_9b8cd6ba0a835eb2da22bb93,HLE,google-vendor-reported,accuracy,percent,true,5.2,5.2,other-supplemental,supplemental-source,profile_3b514549614b819522b86adb,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-12b-card,Gemma 4 12B IT official model card,https://huggingface.co/google/gemma-4-12B-it
google/gemma-4-12b-it,Gemma 4 12B IT,Google,open,88,gpqa-diamond,GPQA Diamond,science,67c3d42930ed74602bce0f3e6aeb82a997165c83a9f680222b27e4071b4da4c8,8263e40bd6175aec016ada4a40dd3693cdc0838024773b66d3360b42b260f37b,8bb642c88acb255828e2b9b1f8de45d7378e2b4db0e5af2b1f5cc9c54331afa8,release_fa97cbc1bde3163225d15751,GPQA Diamond,google-vendor-reported,accuracy,percent,true,78.8,78.8,other-supplemental,supplemental-source,profile_3b514549614b819522b86adb,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-12b-card,Gemma 4 12B IT official model card,https://huggingface.co/google/gemma-4-12B-it
google/gemma-4-12b-it,Gemma 4 12B IT,Google,open,88,mmlu-pro,MMLU-Pro,knowledge,77500c915baa81e8fac0378c8847e20f2312c0a92e6be77494f87723f84d56b3,15d413232dacf9e401f359fb25b4666e595a7d36996e8ef531621ecb52c1cd3e,701b19bfac3ec0298ae4580dfb7a13feef9ff9e77cd0cb2cdc7e0a0e8c25976c,release_cbef64e634e73d9217bfe565,MMLU-Pro,google-vendor-reported,accuracy,percent,true,77.2,77.2,other-supplemental,supplemental-source,profile_3b514549614b819522b86adb,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-12b-card,Gemma 4 12B IT official model card,https://huggingface.co/google/gemma-4-12B-it
google/gemma-4-12b-it,Gemma 4 12B IT,Google,open,88,mmmu-pro,MMMU-Pro,multimodal,8c846ddbb6c8348543ac851b4b044db6ddbe2f58983bc494b688f2d62e321aad,8c4561873e44e627448e353e0e22d3e9652c4cdbd34908950291d7878e414f90,bb9f70a6a5495741815c8a649fc71037491b19c339ce5782e0a65c139014c769,release_6f3fd0029277bfcad78a10b3,MMMU-Pro,google-vendor-reported,accuracy,percent,true,69.1,69.1,other-supplemental,supplemental-source,profile_3b514549614b819522b86adb,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-12b-card,Gemma 4 12B IT official model card,https://huggingface.co/google/gemma-4-12B-it
google/gemma-4-12b-it,Gemma 4 12B IT,Google,open,88,livecodebench,LiveCodeBench,coding,cc47b23796a94a95e343fefcfcb69a0d5b2aa87ffdc24c765e18a6fc3456447d,56e72150ed4cc6072c9711350da5acb66bfe7da1b075cd89d887fa7da483cb33,0b730d9b32134e49951ef8ed8d71eb2e3dae7bc469983e225238f4cef7d267d9,release_1d1cf16bb310d2fb0576bb02,v6,google-vendor-reported,pass@1,percent,true,72.0,72.0,other-supplemental,supplemental-source,profile_3b514549614b819522b86adb,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-12b-card,Gemma 4 12B IT official model card,https://huggingface.co/google/gemma-4-12B-it
google/gemma-4-26b-a4b-it,Gemma 4 26B-A4B IT,Google,open,89,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,4cfc9930b1d7c404ed6d5762b3f3af2f5228e546e778ace5a674dcbd593bff87,02d5ef20549391d85a4ec163bbee89739ec5c80cf2bc5017739c3f0e1bee36a9,06ac0b5ca9aa6c39ffe98cb437a39fcfcce2acb7262c96bbdc1e654d9cc4bbc0,release_9b8cd6ba0a835eb2da22bb93,HLE,google-vendor-reported,accuracy,percent,true,8.7,8.7,other-supplemental,supplemental-source,profile_6f7c719dfe08806351b5e839,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-26b-a4b-card,Gemma 4 26B A4B IT official model card,https://huggingface.co/google/gemma-4-26B-A4B-it
google/gemma-4-26b-a4b-it,Gemma 4 26B-A4B IT,Google,open,89,gpqa-diamond,GPQA Diamond,science,691306ae3fab23ff0110df4ca9c62a279770ef07c7aa05e2084285075d6f5661,612e2c5282a027b331e7612ed465e8c70f185d0a0c66c8567af1365d005f670c,a1c40e0218605478104c85e1c23125c3f8ab74131ce10b070928d9226d484a23,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,73.23232323232324,73.2,other-supplemental,supplemental-source,profile_bc8fb7aacaacf9a5181ca317,minimal,minimal,true,fixed_effort,,false,2026-08-06T23:59:25.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
google/gemma-4-26b-a4b-it,Gemma 4 26B-A4B IT,Google,open,89,mmlu-pro,MMLU-Pro,knowledge,7957c91bbd25f5eef424f8f3afce09cc57a5062ba5d640e1681cb4f3579a8360,84b4cfcca39babd5b69aa7a9cb6ac3b160709d817affe063345ec7428ffe3a58,a2a0026bdf12f6d0b107ae36753b679cab04abadefa418ed0920e49d04843f0f,release_cbef64e634e73d9217bfe565,MMLU-Pro,google-vendor-reported,accuracy,percent,true,82.6,82.6,other-supplemental,supplemental-source,profile_6f7c719dfe08806351b5e839,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-26b-a4b-card,Gemma 4 26B A4B IT official model card,https://huggingface.co/google/gemma-4-26B-A4B-it
google/gemma-4-26b-a4b-it,Gemma 4 26B-A4B IT,Google,open,89,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,956a7d7866d03959fb18dc1538b8607543796abb71c6148ed57da4838435e908,984822f592f1958ecb67ab85147675534db13c363a83b8802077f37e3690999b,e34b79d4bd0fd4a67e6ab3ac50f1df288ff091726118fba383eb00ed5ec197be,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,82.22222222222221,82.2,exact,canonical-source,profile_bc8fb7aacaacf9a5181ca317,minimal,minimal,true,fixed_effort,,false,2026-08-28T12:55:56.000Z,fresh,64,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
google/gemma-4-26b-a4b-it,Gemma 4 26B-A4B IT,Google,open,89,mmmu-pro,MMMU-Pro,multimodal,689a8540de8f3c7a5f0674146ba0f31668a064c5903ecaca16d601bc0d6cf1ed,149ac26e99ff80e5f298c6f184758c11b986fc881eb925310bf9df892ee98226,6d01ec25a9572f1b2135d4ad993b46c30e87e32cbb3c89a35cf3c019bba2bd3d,release_6f3fd0029277bfcad78a10b3,MMMU-Pro,google-vendor-reported,accuracy,percent,true,73.8,73.8,other-supplemental,supplemental-source,profile_6f7c719dfe08806351b5e839,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-26b-a4b-card,Gemma 4 26B A4B IT official model card,https://huggingface.co/google/gemma-4-26B-A4B-it
google/gemma-4-26b-a4b-it,Gemma 4 26B-A4B IT,Google,open,89,ifbench,IFBench,instruction-following,bdcef8a4037ad26843447cdb8142c10cfd48493a506b85d5647dd89769baf493,5ebf3111b8bd70382c9189a6fc80a52c24cad87741579f76b3f7437c100b588c,f55d1bb34adc6d2f246c79d0307e2ab6ee1679c31c3304e9628c0cdc7e169f71,release_4adf348ac2f76539ed7ccb10,swallow-evaluation-instruct,ifbench-test-300-prompts-344-instructions,inst_level_strict_acc,percent,true,72.7,72.7,exact,canonical-source,profile_15d0ef6dad8ff8c425a50e57,adaptive,adaptive,true,adaptive,,false,2026-09-30T18:53:19Z,fresh,31,2,B,2,B,0,A+,official-leaderboard,6,official-leaderboard,swallow-llm-leaderboard,Swallow LLM Leaderboard,https://raw.githubusercontent.com/swallow-llm/leaderboard/c58a13a852eba061d639af7caf4ddbdff33b36e3/swallow_leaderboard_post_en.js
google/gemma-4-26b-a4b-it,Gemma 4 26B-A4B IT,Google,open,89,livecodebench,LiveCodeBench,coding,a45a3ce9fb128956b56f5cf7a857ed169bd7c6bb501a49150c87e22eb924dbbc,13120c5ba16380749a067f7ac080a75d046a91306a339c88986d5c0d14a61c51,5d31469ed4451daa67379e297cddd9f64424f87add212e1bf0b12450d9b3b79f,release_1d1cf16bb310d2fb0576bb02,v6,google-vendor-reported,pass@1,percent,true,77.1,77.1,other-supplemental,supplemental-source,profile_6f7c719dfe08806351b5e839,reported,unknown,,unknown,,false,2026-07-20,fresh,104,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,gemma4-26b-a4b-card,Gemma 4 26B A4B IT official model card,https://huggingface.co/google/gemma-4-26B-A4B-it
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,epoch-eci,Epoch Capabilities Index,capability-index,c5942042a01f40752b051be7090a67d65320e291309887e606d7f6aa80c43148,63b87fafceec035fad6c1b0a5f3e13b0ff369db22d666fa3289e57460b2018d7,f21e6df7bf1044455256ced8c3ed6d506e3759d6ae79a744c9e5420acb8698ab,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,150.46,150.46,exact,canonical-source,profile_16db478ee7ae220f4d066083,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,aa-intelligence-index,Artificial Analysis Intelligence Index,capability-index,c3de7b8ca1a567b7398e148f6574c0a09edc508a66f48fa8bfe954c7dc320e9a,bd2c26a05279b976466bd63beb59de5cd400dafb3e6c0e44dad20bd7a88778f0,2ed15744d2f984a906bfd4d45bafa9ea8325573bcdb72a803752953a712d5de1,release_c1cd22cc93bbb595b7e3c6f3,Artificial Analysis Intelligence Index v5.0,intelligence-index-v5-0,index_score,points,true,12.910162,12.9,exact,canonical-source,profile_20eb51e9f88cfd04e5ce2e9d,reported,unknown,,unknown,,false,2026-10-29T10:53:25Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,arena-text,Arena Text,preference,46537e998deb3080e7bf79bfe42035e02a328311c3a73cdf600c1eb84f5a429d,5f4430875bade41b3cd464e25813a44ac8ad20889db4985f572fe36052e6b67b,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1405.5220424213123,1406,exact,canonical-source,profile_9afb89ae1ce26b5639d73387,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,09c331e89f7d2cbc19db734f2117734ee334cad0c22742500cdb3d9816de6769,1994683afa46673c48d1181a83f144d13047321b137011a89f26e26010b7ce2e,ace1ba8ee6ddec70890eafd20d98cb24e8cc615fe1b0f5899e8549c87453e562,release_fa9e908ac04b067eb00182d4,HLE official leaderboard,published-leaderboard,accuracy,percent,true,31.6,31.6,other-supplemental,supplemental-source,profile_d2b3708222c5ac4dd0dbf632,reported,unknown,,unknown,,false,2026-10-23T21:21:02Z,fresh,8,1,A,1,A,5,C,self-reported,6,self-reported,hle,Humanity's Last Exam,https://huggingface.co/api/datasets/cais/hle/leaderboard?limit=200
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,cfa1def513b2409506a4d7346fab88b5e5bc36db651ddae62ff506993cba9376,8e71f13fd1a0d250249e9b3074c5d25d38f637d9a71e4ec7a7039b70965c0729,1b28dab1a1a430aa0e5ed2ea185cde86e3d2be3b91d72a14f3efd94953e40fca,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,33.317887,33.3,exact,canonical-source,profile_20eb51e9f88cfd04e5ce2e9d,reported,unknown,,unknown,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,simpleqa-verified,SimpleQA Verified,factuality,f251d79a001dfa252f7d162fe4d54450457740baf4fb28ee338cd6321be10d9d,e4b4b55b1c4a8787dd0ed9c78c2cf1a52cae3b719aba48e1ff7438ca1b10c6f0,233a1f4e148821f06ce49380fe8bc1814d008506ed8870d14305ed38a37e8dc6,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,19.1,19.1,exact,canonical-source,profile_11897b7e05ae3b489552bf1b,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T19:36:17.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,gpqa-diamond,GPQA Diamond,science,710291b8cf7860f9d4610e1658c5359abb8dc7f7f24211f21ab83c220e4ec962,0835a475ee249273c8e4a92554f5995d1fad29fbedb197e29264945b021bff5d,b4e40cd174a8badf46e20b26e1810dd03fb5514adfbe5d2c780c768418782efa,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,83.586,83.59,exact,canonical-source,profile_e7221fcbe0df2d3f97b361b4,0.99,unknown,,unknown,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,mmlu-pro,MMLU-Pro,knowledge,dbfdf6fa117988b7305e64080f13896b0d47ad8b1c0262ff7e649d1d3d64e2c1,09a4ddd5c6cb9b6f9e8210c233c2b836000e54d060cf3d2af70efd5319f2194d,79eb70b78262259dbf59d93e79b296bd094bd31d0f8cb5dae3362b0f394979f4,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,85.566,85.6,exact,canonical-source,profile_4afb8d94c07f2a4577e683fa,0.99,unknown,,unknown,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,862f340b771080696b2bb00611bdd0885dff9b3706e26104beff2c3c79715d96,d0ca8f22660cd4586e2ecf98a914d737760bc1f9b9fd5325459f8a8dd5a24dfd,5ba7da82e26519352aaaa2731865202ebb0ad36f685e72035b623886ce822905,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,90.0,90.0,exact,canonical-source,profile_11897b7e05ae3b489552bf1b,xhigh,xhigh,true,fixed_effort,,false,2026-08-14T14:19:47.000Z,fresh,78,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,frontiermath,FrontierMath T1–3,mathematics,3c55539942bd956465ab01ea93ef2733e89b60d6d52b4a5f6463b95a06128279,acdf7419e13b15645bb5757240ae8785217a1870730cfc1ae4473338dc4caf1e,1175a744640d4a5c8c5837b22e5668b81b7454a29aca4cbd40a878a02f6aff48,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,46.31578947368421,46.3,exact,canonical-source,profile_11897b7e05ae3b489552bf1b,xhigh,xhigh,true,fixed_effort,,false,2026-08-14T19:59:09.000Z,fresh,78,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,arc-agi-2,ARC-AGI-2,abstraction,fd2ffa704695bec3ee000edc6a070e245993d98147542773ca8b77127aecf039,6a3fdbb035c6f4bdfc703192088ebf4543574902a5083f0c6b9de6ddba89b303,f1c08ad01e5310c4d0aa22562f56e8ff60f51fd82895da2534adea920bcc71c5,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,40.138888888888886,40,exact,canonical-source,profile_11897b7e05ae3b489552bf1b,xhigh,xhigh,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,frontiermath-tier-4,FrontierMath T4,mathematics,9360b80d988e91b0bd6e9b89ea36598c26cdec3828bf4602e1f9d31b9a87c845,7f1b1e9202b27f97d0386b43279e377df953efb8f15c9d1db91897d24ae0ab4d,8a4e42c5a1b8e6bf5f6c5dbb9c7133e83186afc175e7a112a74935b3030c4f51,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,17.073170731707318,17.1,exact,canonical-source,profile_11897b7e05ae3b489552bf1b,xhigh,xhigh,true,fixed_effort,,false,2026-08-14T19:59:09.000Z,fresh,78,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,mmmu-pro,MMMU-Pro,multimodal,d8aef58a30e3f8647d08b4a09eefcd5e5816aacc2f2c8623d130accd3a44ad79,c3c12678fe5c75076eeaa61f51888210044a511312decbb19f0f22aba63369de,fe7a06af4072ada5e73d091112397576edd5b505c7eb7f17e938be8883165d4b,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,74.046243,74.0,other-supplemental,supplemental-source,profile_20eb51e9f88cfd04e5ce2e9d,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,critpt,CritPt,science,24a8fdb723dde086ac7598dc35681f44cd920039422bcd93c9a63c77a241d569,944d910124e7e3966227d67d0b7928e262c858d292c86a91427ae12b4e250e75,057f0d50f46701d2e33e77c0d7fb81654a0fc92fe0ab67133a18d8fa4479c4af,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,8.285714,8.3,exact,canonical-source,profile_20eb51e9f88cfd04e5ce2e9d,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,arc-agi-1,ARC-AGI-1,abstraction,55a43d347baba0e8b66db0de0bdd8589df35592f4b8b6a08ef734cf2b57ac89f,f991f89d449f430276429e1c89defe5c7fb989c747084d33a287642ca7de73da,b23c7b89a7b59a9f38b93fdb7965b185db7bf5d22ddf0951429c283f83756fae,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,84.0,84.0,exact,canonical-source,profile_11897b7e05ae3b489552bf1b,xhigh,xhigh,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,swe-bench-verified,SWE-bench Verified,coding-agent,3b77ce662616620e9c6a032a0724d50774d7095f5f810955a6db9182ebec2808,b26533c2ec49f018c04fba36a7e6ced76b5293a09f63c6dd4603d9f10cf88850,fab4fc70e27554b8de56bb649d0feb76a74eeb75cda84a66163688c988a397eb,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,82.2,82.2,exact,canonical-source,profile_703781bfe8a825dd1e4f60a3,0.99,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,ioi,IOI,coding,fed8b5a8afe3c669d323e9c4560745cc0498fb05c0502fff490c8863a7e3e9fe,70d3c0bbc193bd96c09a62150878283b6f1a5f36c7121918d05e95b50e2f4647,c7fc7ea5ca742d26b71ac37e76c415e45b367b43a4fd4a342a4cf757b70c7d73,release_e9a397f5dc7f7ff80be0a322,Vals IOI v2,vals-ioi-2024-2026-overall,score,percent,true,9.333,9.3,exact,canonical-source,profile_dd301f51b56b79061d8b85c8,0.99,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-ioi,Vals AI · IOI,https://www.vals.ai/benchmarks/ioi
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,livecodebench,LiveCodeBench,coding,29bbb351dda5ce6409f95629668a151af52e580186616a8b075a659c24b0df9d,af8d5cc203876c3e8e99591aea80c5986d9c4e477425c6630324e95030a8f54a,5908a331e7fff5448ab4dfd47c9906598a798b61151b8726e4d43e0ee056b7a2,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,85.931,85.9,exact,canonical-source,profile_e7bf88fa29d99292354422f8,0.99,unknown,,unknown,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,scicode,SciCode,coding,bf9dab9475f01d92f0b20b6a1dad3c3892ec68420f80fd1937e760e2de836f57,94649923025f12a9518edfe452268738fce0ec650aa62ff330720f6ee683e473,f0903fc29e56ecd3275b78d6d34d7b354f54d7774c9a741193679cf523544919,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,49.6527777777778,49.7,exact,canonical-source,profile_5ef7626ca399657d16aa77c6,reported,unknown,,unknown,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,vibe-code-bench,Vibe Code Bench,coding-agent,683d4ec649b5c8dd59da94c237715465485f86464acf1ea76b19d7f3d37bb0b9,d023213ecf3ac148545daa0c9ee76728e822d61747daa7c5d07e2ca429ed13dc,25d75f6996e99bf80763cf17a190c08520df25c0829c3695e4662c128b073620,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,19.052,19.1,exact,canonical-source,profile_1a5020ecdcfbe5f23617785d,0.99,unknown,,unknown,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,webdev-arena,Text Arena · Coding,coding,706acc7a1074d29510952fbb2a57d999238f767ac2a5cd0a4b05ecaee36cd9ab,6be81d75ca36c070685854ab414620cfc24b1cf82e83075f994e33cf24be9e46,7bc13f86b676cd657a6a1bbe561c8fd3d20a00d987266c3c3fe6c424a8d93172,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1409.1376517895078,1409,exact,canonical-source,profile_13830015e96667e4be7dbbf9,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,programbench,ProgramBench,coding-agent,262fb50bf303095582c9651b5027714a318cec1fe40e0e2a426c207619b2236b,2072f473a5597048b3ceb70726364323057c9b9649d23bbeac4c9df3106d0470,dbe63254234b16431309b2a726973637f4b012dc4e9cf08b25d861965d3deb2e,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.5,0.5,exact,canonical-source,profile_a4f343eeab832539cf337101,0.99,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,9904a4950829dd35162b59c6e7dbf77d498879279e3a2839e1575777af89eb30,0dba9f1ce633c6619848b20894567a8d831e6022a6bfe1ccf44e2a7a98de7229,50d5b67ce1f30bb532f62334f5c2967061a39634a3df5e532823bd42bd7801af,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,55.056,55.1,exact,canonical-source,profile_629f3e62b174a633d2359fb4,0.99,unknown,,unknown,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,9a42581b5d16ecb7556855d8a550104ba89dd1794ebcd49de191e57f7aac2cd2,f72cc14e10264ad2d804d4a6cfa44d5259906a43266a9b7454998ee16b7e3f55,5807ef6f33eab5251c77fc6b4c30f3bb7faf37df4949a812ec4889ac489272d7,release_d944507cc6b610fbbf6eb832,Vals Terminal-Bench 4.0 v4.0,vals-terminal-bench-4-overall,accuracy,percent,true,0.0,0.0,exact,canonical-source,profile_71890960f4f5938a7741e052,0.99,unknown,,unknown,,false,2026-10-21T23:21:14Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-4,Vals AI · Terminal-Bench 4.0,https://www.vals.ai/benchmarks/terminal-bench-4
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,bfa4d297b930b6b6f00d2421b98a91be6bc0fd91564543c2e357e7bea94ed9f1,de3d5c152a0f163f7b25700294fa1628c7af66ea8de67d4edd37eb988c0e9f5f,6bca21909196ef9b1bc2254f2a5b35d7288ac76ea0b8682752a2e683e92a3520,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,18.762887,18.8,exact,canonical-source,profile_20eb51e9f88cfd04e5ce2e9d,reported,unknown,,unknown,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,mcp-atlas,MCP-Atlas · Scale,tool-use,c660148e115df15ac70c9c4d89a5434b72d0154b808ceb1bfd5508054c5a3ec6,cd07292fe92ae68cb27046d6b2f8c50274c8250dc5dbefebc39c949a7283393d,a031587fda1161b928eae2d5b5f1d93e326ef3faa6b1d4335ba056f467f8c5f1,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,79.2,79.2,exact,canonical-source,profile_318b9d51698a9143a85074a8,reported,unknown,,unknown,,false,2026-07-30T19:17:22Z,fresh,93,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,arena-agent,LMArena Agent Arena,multi-turn-agent,7fe834bae68a47013070f025b893518e5be28b22f49878e27e4060a64bdf404b,1f5c234b174492be761d1998ea7ec317606b7d1d6ff656ee0b4ecce8a673a130,b2a02e9a9bd2c163c3cd1fea8cf66291a9f8b24cc004173e5ab19155e294c0be,release_7d8f719478c55b93fbf0585e,2026-10-28,overall-ips-aggregate,ips_score,ips,true,-0.17010177604397714,'-0.1701,exact,source-specific,profile_9487927c8b4ab32b43e1f141,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-agent,LMArena Agent Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/agent/latest-00000-of-00001.parquet?download=true
thinking-machines/inkling-small,Inkling Small,Thinking Machines Lab,open,90,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,2884f07c2d98d5fe0baca1d650b351137dfdb4c7daac4cd6bb2d3fff30de51fc,8e7b220e390d5e52cdd7a4c00e0e5598a80403f93b394626a99a924355610c68,160924b0df04a32838bf5f5b4bb87eaa317a9eeeacfae44f8336c34ec4be552c,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1132.56,1133,exact,canonical-source,profile_20eb51e9f88cfd04e5ce2e9d,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,epoch-eci,Epoch Capabilities Index,capability-index,e3e3b19e2b995b957fccc7df30da2669fb0fc5beb4593ebb70b453eb0d12e854,57784bd29b5f315ddf4b625fb59b7facebdc30ecdf55c58338dfd90ead8f1e62,e57dd7cbb950e5797f9f0fe81ab5be952e8ecdb43f3bb637edc01d37a63a3959,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,148.11,148.11,exact,canonical-source,profile_319ef04e6ebcddf6677b39cd,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,arena-text,Arena Text,preference,6502ba4562588d3ab2ff7e3de5d356471ed46f0c7dfe61ae977a46971198e9e0,95e65031829995a84133ca0f058f6375a7839b3317f6dba276f7ff1a1845dbea,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1449.746776109821,1450,exact,canonical-source,profile_6ae6ce92018cc9651626b3ec,adaptive,adaptive,true,adaptive,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,08d0c8af8379324e485bbfdd5575e10bca95ff24b78181211f21ce91cb557da0,6f5e6236af2b155103664d8bdf530700e0a01da4aea73822e28fad5544d52573,6a88ff33bf8496a3f701879cc798199c8b15ca6088d9005645b3401d2082fd7b,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,24.37,24.37,exact,canonical-source,profile_a22fc3d76219aa5ad672fcbb,reported,unknown,,unknown,,false,2026-02-13T14:32:29Z,stale,260,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,simpleqa-verified,SimpleQA Verified,factuality,c50f4133e34cf346498948b130308fc8022b86ed1e86af1548e9952a9fbf851b,f8a3858077fab309f8086e27150189245ffbaf7067eac5c3e766a7ce7bd9d473,9b27a207e2a514be653bf3478a54786b2e28f6a37c4d6332e9e947a6618acd03,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,34.300000000000004,34.3,exact,canonical-source,profile_7aaddd2f0cc6c1a8ff26df64,reported,unknown,,unknown,,false,2026-08-27T19:35:45.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,gpqa-diamond,GPQA Diamond,science,fc648ecf67ee1be770a0f96cb74df0105bb74cad8cfd30ae3c9d7068ed7f31dc,9fa688da9f8f5ed3b6b6c02894d00e6b664e7290b51fc2a5ff6ac09828b43ad9,423d01a485c271876c1915edbf0abe3ca359455ecc1eddebed52ea6de97eb299,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,84.091,84.09,exact,canonical-source,profile_6ae6ce92018cc9651626b3ec,adaptive,adaptive,true,adaptive,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,mmlu-pro,MMLU-Pro,knowledge,688702f41eda54165b2e871f54106b358a0c601953e7338623dcebd5bd940f15,1a3b39cf7c97fc9328d3297f47f2cfe9a2dd3916845523b1aad659d9e55d6dd5,97e9a5836bd57f0f1eba1091596dd1c3a3974f9f73de6bf4cbd43fd15c1ca3c3,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,85.914,85.9,exact,canonical-source,profile_6ae6ce92018cc9651626b3ec,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,919b3cea001418242d59467bddb96b27f516debd934379ac16f1351816ee3b6a,ed9a5daceebca5335ac9cf309bc8921076e7be01636f9efe832a13ce6a42395c,53b5afd59ed2fcc1bea62c05d6126c9965d6161c4095ab8a5398b2b254a04a9f,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,92.2,92.2,exact,canonical-source,profile_04028ed53a8f5ae82875593a,reported,unknown,,unknown,,false,2026-02-02T10:53:14.986Z,stale,271,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,arc-agi-2,ARC-AGI-2,abstraction,de92e702186834573601a58dd39024401df3f2e3b2784efaf13ae9bed02e6f3d,ebcd5666ae1d0c53b61aba5581361536f38f768b402706ae48edc21a3faba355,d2f5a5a3807c2d47e07c18d7540fe3125ac78f0374c963dd93a0a538bb54bb38,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,11.81,12,exact,canonical-source,profile_5231d77a85df897c20d6ce6e,reported,unknown,,unknown,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,mmmu-pro,MMMU-Pro,multimodal,4ed87390b66f6b9d630ef32cc517ae9c792d9456c661c0e4751243ac235b8da0,131e449c81fe0ac9f4246d63c2b3ed434b11a20f6c0c5c3b9f01cd1274c9cf89,9143580b75a90983f666093f3cbc153429479f600899d32e509199afd95eb35a,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,84.335,84.3,exact,canonical-source,profile_6ae6ce92018cc9651626b3ec,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,critpt,CritPt,science,424f5358b9cd934b7441a5a75f8b0f66f1c2e2c94c41331a6341899ca1d9f0ee,9b71e82ddfc4af314a95e159eb08a0d73fc8eb6abb6d1f2ae211e4ba697c88b0,d54638f4ca45fb6c067f5ce4c9ba2dd1e3cfac45b51ddfe7af5014674c9ecbb5,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,3.14285714285714,3.1,exact,canonical-source,profile_3274c202f27b162a812bd5b0,reported,unknown,,unknown,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,simplebench,SimpleBench,reasoning,ba76bb771a11466d19a080d2d32495484f5a3190ff1e79306136a6d96f00c537,652867cc1c8aa4baf1b9ba1cd58cda5005b8195d16e081b0c6df3ea7361241a1,56f2ca3e622b4d4d099aec16d36c9c43e7e45524034c416dc2bb838c50d8990e,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,46.8,46.8,exact,canonical-source,profile_cdd77eeb27e64fe682104bb2,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,arc-agi-1,ARC-AGI-1,abstraction,bce36270835b9e565ec1f01f90a613cc3a44e0b4e2651667f6870e83ee7f388a,dd7b161deb464e517c56fe5b5e048573b9a1b3fe075fc1fdce358ef8a905ebd2,cc43816fd6d881f3b38fb823a89342968180109a116a45d9f4de815ec4d9cf15,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,65.33,65.3,exact,canonical-source,profile_a605b31119223a617f1c6466,reported,unknown,,unknown,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,swe-bench-verified,SWE-bench Verified,coding-agent,72522dd24c6d25ec39a4dfe4309f6ad90ed7ce6949081a535584e925af757126,c1231cae401dd9744d098e470bdeba618e54c73e83d1275fb12a804cfc0d25ef,3aeeedd7aaf29a157ada175202d53fa3db6e75153173f3a290265a0bc18cb6eb,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,70.0,70.0,exact,canonical-source,profile_6ae6ce92018cc9651626b3ec,adaptive,adaptive,true,adaptive,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,swe-rebench-v2,SWE-rebench v2,coding-agent,4d9ea716c84dbaaa7a43d5a830adfccf8586f5538ea064bcd73c7464cb7e5630,5ebda0432145016aaefc28405cad973ad75bfa27af6115b9d4d89ca511b2c057,141d7a694df4e1ca70e967074cda149ad778755d6ce8798e08f8a094860fec90,release_2125f5e1cd1e1b1aa6168fb5,rolling-v2,2026-02-01/2026-03-01,resolved_rate,percent,true,58.452380952380956,58.45,other-supplemental,supplemental-prior-window,profile_025533521cd219b3ae285dd7,reported,unknown,,unknown,,false,2026-03-01T00:00:00Z,stale,245,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,livecodebench,LiveCodeBench,coding,992b37e67975297b06468c6e9ce6f766ccf10eaf5742fdb8c0e1269b6b082b90,b6fb3f0f2d85630c92051c15505df2e7f468c7e5a275fa248d90758ada0e6806,2af08ef4db534aedceaad55ef487e6530a8d7db766b9491fc248b5d9cbd288a7,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,83.868,83.9,exact,canonical-source,profile_6ae6ce92018cc9651626b3ec,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,scicode,SciCode,coding,a40a91379dca063e09e6afa1d513f8aebc0c0d4c520cce77fa8b22ed69319f36,579c7d0908c30a1369ec3f2e20e833bcf0f4feef7f56d7372e3541d00d936d56,9a751ce3717bf2e8fcd5e6795198ab13ebf865224c830224359510696dd32fa7,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,48.9583333333333,49.0,exact,canonical-source,profile_1307685283e08685791dd3b9,reported,unknown,,unknown,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,ale-bench,ALE-Bench,coding,5e669305c42d4e329531a630504c3bb2a9e777b6754ea185e9cbc2a934aad873,531f54234feb8424c7d322ea443e1c81b2a1a1f3b65152ab6e2de9c3c173159f,860d50f7d2ce508d6a774bf679003e9580848eac4e86b958ad49d04a68a96675,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,821.65,821.65,exact,canonical-source,profile_4a115fb2a00cd4c6610fdf9c,reported,unknown,,unknown,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,vibe-code-bench,Vibe Code Bench,coding-agent,f589efc0e39e68b102cf22e0070190bca353284320635917d4aa5a1e44834db9,1b475729c12a6f6acd36647318b8c478dfa67352e47e075875a8cc5e04cb32dd,25b79921a5da4317f38186d40d9c1876794606d48cab970584ec97391d864284,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,17.536,17.5,exact,canonical-source,profile_6ae6ce92018cc9651626b3ec,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,webdev-arena,Text Arena · Coding,coding,efbd786303f6ebd8d02c01f8faf3973c49fbae4a773d311fad2b27b321df4c10,b2f49ba2077ca3f96f045d25772fee110bcefc191b71d78d9ce8431745351d1b,919b8d94f2472f2879c56026df59729b35ab43c62f1c3e55773886879416c02d,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1436.2450288272603,1436,exact,canonical-source,profile_6ae6ce92018cc9651626b3ec,adaptive,adaptive,true,adaptive,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,e9e1b1875a980a50790aef4b7bd3a16816d7485ebddbb99314e79934d41d738b,07974584d555a6f1248a9c191677f54a12662acc57cec0657f81819e8ae78a8f,05d83f9558f680b0715f571eafc49a45ac79be2670a2c6021b6d03ba4920b20f,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,41.948,41.9,exact,canonical-source,profile_6ae6ce92018cc9651626b3ec,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,vending-bench-2,Vending-Bench 2,multi-turn-agent,bebed54504a4ec3a55ef290bbbed321961f946604ffecf82de8bead4a97c7b70,008d81ed10abf19df9bf1d6c087c6d8709c217540b467f46cea6cb31cf143201,ceb47379ebb944853a74277a42f187b11105aff2cab403497864f58f2829b934,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,1198.4579999999999,1198.46,exact,canonical-source,profile_c40dc6af044b94a7c275bc96,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,df62ac856b75803896ee538cec1ecf384b59e37271a0cb66cc0d93b521e313c2,e293713c57826e5cd3a1e21a04422b14dcca8e5a7960b865f3a334f8c650c74d,a67ad4ed32c6284b0f077084f97c8e710bb6286bc434a416e8bde92edfac81cf,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,846.64,847,exact,canonical-source,profile_6ae6ce92018cc9651626b3ec,adaptive,adaptive,true,adaptive,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
moonshot/kimi-k2-5,Kimi K2.5,Moonshot AI,open,91,weirdml,WeirdML,coding,c9e50fffd268058791314f53e802a757246f18714f364d5a730016e4f978fc47,2599b340f53510b657e070b78c5518f17199a0fa6c3955657ed9afef5f17ff0e,944c42581b9cf19afddbc0b18f843bea0572518f388d8eba131a1bd77f675dfb,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,45.6,45.6,exact,canonical-source,profile_4ba684af8be038b2c7303841,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,epoch-eci,Epoch Capabilities Index,capability-index,6ff654ab1e336d3d61a2860b9b47f915b7793339bf8c830a571f74dd435bd195,14fe32bddda040186c98a1220db3648ef61c257219141c6123d7947fa507833a,0522a423c124834e0a23595e46e5eb07c3d85add8e954dc06c03db7ffe976938,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,150.23,150.23,exact,canonical-source,profile_985db51e44c0b3d3bb9fbbfb,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,livebench,LiveBench,general,eb8833b05acd587662c89ac13779200cd8a0c61e1fa742b6a9f1dcd5ae6aca82,0a86bf6a0b6b2c7a592bd30688ec9cb5b78e0941279851c764600df7d40ac5b0,a84af31207ac7d6490dcb1124920c18a5cc4c38ca783ab2559a81357f257ddaa,release_c94a1cfb9d281df4254ff243,2026-06-25,overall-seven-category,category_balanced_average,percent,true,68.41172619047619,68.41,exact,canonical-source,profile_47aaeb01147213e8437e7340,reported,unknown,,unknown,,false,2026-06-25,fresh,129,1,A,1,A,0,A+,official-leaderboard,5,official-leaderboard,livebench,LiveBench,https://livebench.ai/v1/table_2026_06_25.csv
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,49f602ea5c85b8b9c26e8168c2df3984f78044eb902f4d5db437d2b2fb6ff421,23e9fc12a95af09be475898d6e0747a08ad723cfccb473658ef550f642691065,cbf12bdea6cf080e4b48675595fb2285a324557cf6454f20733ca656213d4980,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,35.032437,35.0,exact,canonical-source,profile_c418354814777762766b7183,reported,unknown,,unknown,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,simpleqa-verified,SimpleQA Verified,factuality,38381dc8d16af483aa95c85aea32f80cc6fd1b4a57b6c506db85840f210160d8,5520a38af43a5b177f3b4b1606398c8892cc57722cf57cf3bc9bda2ef93c59a5,aa2afddb6862878a83c16956406d08720ef942d16c7bcb30f7d35c9d8326149a,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,36.5,36.5,exact,canonical-source,profile_63aaba2756ec14ee77fae61c,reported,unknown,,unknown,,false,2026-08-27T19:36:03.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,gpqa-diamond,GPQA Diamond,science,81ce2acaad30b48c8e7e6773e970285575b59cf1cfb050f146e42662ea4990c1,1beaec7bbdb4beace853476b82f8f395ea17f0f1000b15b57ebd042133165cc1,31c9e1853f57d6597a92c38c93fe9f8b762953a9f551ebb99438326d6adaf99c,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,89.59596,89.6,other-supplemental,supplemental-source,profile_c418354814777762766b7183,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,ffe88e2ae9dea362b4b9af43b4b14fa9ea12c0e61d03b3103f48afbdb2d54f3e,fdee7e95be1875bc6af041247fae7aa09d68f462caa02d76103285318d7f1d0e,ba1560a0e78954e95a1fbeff6bb7dc8f519b825a9793b92f0c72e4c4343ebbcb,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,95.55555555555556,95.6,exact,canonical-source,profile_a87929f64e446284534a7d50,reported,unknown,,unknown,,false,2026-08-07T00:20:43.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,frontiermath,FrontierMath T1–3,mathematics,0d332f6735a6b86f3d2f7ed09f6daa7e0bbdf2f7363fa6e9571c616bfaccc550,943833210192902bdf3bb3e20b09f2c02a31b7a938b519a489aa395b9aa68023,c934cab44e6163fc127623626f7cab37035e2979a6b4784688228eb71f9b820c,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,54.03508771929825,54.0,exact,canonical-source,profile_9137a8c8a7db7d39dbe479e9,reported,unknown,,unknown,,false,2026-06-13T01:39:44.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,frontiermath-tier-4,FrontierMath T4,mathematics,05ade21a00f2daffe49608971919cbb8cb1f5e96d85343958b29ddd4fe897b69,101c2feb34110401f816b734c334dc9aaea3ae25d92baede7d61ec311011e00c,cb655832235294bac51fdb680d75442eac73bd158681fbf20e452ca29f822269,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,12.195121951219512,12.2,exact,canonical-source,profile_37090ab2b47281e7d26ce1a7,reported,unknown,,unknown,,false,2026-06-13T01:39:44.000Z,fresh,141,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,critpt,CritPt,science,c73062a96f3409f114aaae3d8c7842c662176932334102ad1491f383a798a877,868140a783dcea3332e634fc766749a40baa73b9bb67297c8587901ea17bbd89,2528582e78dc47e9aae05644e86a22532b5609720f64082a54e09642fc3a5633,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,10.0,10.0,exact,canonical-source,profile_c418354814777762766b7183,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,simplebench,SimpleBench,reasoning,1cc721ff6d34a396a3a8a8d82a51a7907a0c096d1624e65096d1c9077ce44a2a,484954f7fc29c6141f17b4402127e51b27fd85a3f98a4dbbf848c8f5d3ff085a,bf92dc42aced9cbce2c8fb44c7e2bb56f53e75fb33c035f97e38139ed580cfd4,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,57.9,57.9,exact,canonical-source,profile_8861fafa0de5f1d143a949c9,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,chess-puzzles,Chess Puzzles,games,4e020953e924182ba85a7f06e1a89f5537070480cd016026c5ad56abf67f57b6,264432ba8c0d7a57581d823aa86cf6453d1e74ef916abf5155b2bb52c2bce66a,cfe739535319190b76841d0eb8b10e35fb5ebdff216eb6496017cfdbed8cabb1,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,14.000000000000002,14.0,exact,canonical-source,profile_a5fda2880c2096a50b322ade,reported,unknown,,unknown,,false,2026-10-13T17:13:43.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,swe-bench-verified,SWE-bench Verified,coding-agent,50e6f9a2f250af4b4b76594bea5f77133e379c00eef9c08b2f3aa72799ccda0e,460ba4c48ec551141bdf47f1877d0454db4d058d0ad05933370a997994556970,cf2233bfa2373c71c8ed2087846a2f4d1737f015763e77f6af90d1af8b710e6a,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,78.2,78.2,exact,canonical-source,profile_4563edc5bc06fca8cdd968dd,reported,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,deepswe,DeepSWE,coding-agent,848afee5ef88db978248d9efbcc83b4123d280b45f5665d0648c0c8e0fee7524,c1e8cbed9a32399d6f611ef470ec52cffae1aadc3aeeff2ffd6eceee9ed501d3,779874d29535932d55ae2983cd4c98afc7144b8b4301e0a615e94f526da87de2,release_4731f79a89e61b8ad678b622,DeepSWE v1.1 live leaderboard,113-original-long-horizon-tasks,pass_at_1,percent,true,30.53097345132743,30.5,exact,canonical-source,profile_1d6cf902dce1551711bea365,reported,unknown,,unknown,,false,2026-08-29T23:52:22Z,fresh,63,0,A+,0,A+,0,A+,official-leaderboard,6,official-leaderboard,deepswe,DeepSWE official live leaderboard,https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,frontiercode,FrontierCode 1.1 · Main,coding-agent,aa4ab57ad65ec5b5f894537d5dc45afe5100160109d0ab5ca71b329c92934927,64469ff3d0088423c40720e54568c4ff3eafe37965d82461eb4a0970efb222bc,227f91e8c423af202566cf13b45663433088c2b5a0344b93eabc4c5fd2951896,release_28313b815e5f20ff3646855e,FrontierCode 1.1,main-100-hardest-tasks,mergeability_rubric_mean_at_5,percent,true,30.06,30.1,exact,canonical-source,profile_dd4b35536a13bf97a852f231,off,off,false,none,,false,2026-08-28T22:38:18Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-frontiercode,Epoch AI mirror · FrontierCode 1.1,https://epoch.ai/benchmarks/frontiercode
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,livecodebench,LiveCodeBench,coding,83db7b37e3c4d90be9438d697b3556f37698465e3022dcac047fbdae0c0e78ff,e7af5dd4614dd5c97c5b495ca91a2e261e4f745868a020143c3b7fb27833b2ab,d08af2a5b5ddf37a13d08b258ae499a5bbe2fd41acf262dde3feca8c731c36fe,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,82.05,82.0,exact,canonical-source,profile_d975f32ded4665b2f1e01b66,reported,unknown,,unknown,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,scicode,SciCode,coding,a786ccbd59e0f0ca58f81ef55becfcd1a3f56fb7e0f27c5e866fc5b8bd9a169c,329646bed6c21f04d5d74613e39921ca7db618f7960d7e87add14355a8a74cb5,e8e79e98bd24134c864da2b0e0dfbd0d74a2bdbc0d01c47a64397f02de275a25,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,47.8009259259259,47.8,exact,canonical-source,profile_1c34c4174282cba541b7d9bc,reported,unknown,,unknown,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,ale-bench,ALE-Bench,coding,14642ae646028eede1ab351078b4cc6f987ddc3cedb24ada6d3c583abb71753e,8a250df5e71b7871cef738921f76558072fe71b1d7fd74135d101823745ac660,e08b8bef922d2365af23dc1a66a5964759321622b95ad38de5d62684f7c89410,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,886.23,886.23,exact,canonical-source,profile_623f2c3e7cb74eeabc3b0ea1,reported,unknown,,unknown,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,vibe-code-bench,Vibe Code Bench,coding-agent,5fc1e40f75312af0080e341f5d19ba95fa69fbee7519157cbec0504e23cc96fa,fcaddd0f33c844b034a1e53b6796516d4b18241b9fef15d31618ec82e736a603,3d1932a9d9af66f3d40d79a7c2832fc36a316362897d293c86a16f16042cf41b,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,47.211,47.2,exact,canonical-source,profile_e042f5db1ee3a91dd7a3bc95,reported,unknown,,unknown,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,webdev-arena,Text Arena · Coding,coding,ead56752fd5ee9b987204f4799ced7c765fd2da8b1321813adfd67ef69ac5b92,bfff7ec0ddb824928eaa1eff3f04580efe1cea9863c1ba5ccf8084e12aa22df7,04bc93151a5e84d1a5cbbf6a85efeec55df22ef665fafe007d005255b851ce89,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1473.5355204604502,1474,exact,canonical-source,profile_d8d2349529d194d11d6a5ea0,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,programbench,ProgramBench,coding-agent,1ab3ad9013572e281fece9e46a69dd7a9c121086fafbeaa0cfd3a86d3be0948c,6f043dc4ea237435ec49159fd5b0aaa239818d2e5c543eb556cdf340737a5f13,e3e3464450eff5c115e48dd7d8023c66a76a14e7d37083aecd1cb72d127657a2,release_88931830e68b802f00d3f781,Vals ProgramBench v1,vals-programbench-200-public-tasks,fully_resolved_rate,percent,true,0.0,0.0,exact,canonical-source,profile_3d974e559f7d9d71f38a39cf,reported,unknown,,unknown,,false,2026-10-21T23:21:15Z,fresh,10,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-programbench,Vals AI · ProgramBench,https://www.vals.ai/benchmarks/programbench
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,325bd85ef5e9788ed6641dc07201e68defef06638d04dfb49b4a4c7fa72b01e2,ac3fef2d020fe6e804e072902bc1b2d1e18437257d8a3d66a9deac06c8933718,627971640dbf722ed7d004d14af9b96bb57d0780f5336de49662bdc58f7bf696,release_3a1f88ce28d6c9608c34ad04,Vals Terminal-Bench 2.1 v2.1,vals-terminal-bench-2-1-overall,accuracy,percent,true,67.041,67.0,exact,canonical-source,profile_f7cb373722491bb55787c0f1,reported,unknown,,unknown,,false,2026-09-02T09:40:43Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-terminal-bench-2-1,Vals AI · Terminal-Bench 2.1,https://www.vals.ai/benchmarks/terminal-bench-2-1
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,2015ecc01bdec07e20db77bb8392b711f56f068a7186763f5ba0ad076baf10ab,fd87050f863c05de173eb58a5a71ea23e9f76bde02f43a28021bd2bdad29a833,90c7acd2ed3d0523b4adfd570ea5eacdb5c711cf72f5d905d945974f58aa54a7,release_b7d9ab744ffac0cb856e114d,Terminal-Bench 4.0 (Artificial Analysis),mini-swe-agent-66-tasks-500-steps,pass_at_1,percent,true,1.010101,1.0,other-supplemental,supplemental-source,profile_c418354814777762766b7183,reported,unknown,,unknown,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,3ceb234a7f0c3bf3751f735f30a9a733a5f4e2293369b672b6706389bb3f65a9,a1a697f2c274336eb89c5eeebe4bb16db800d63a5e11f23ef783cfe5dd9f67f1,d80d07f36bfd41d5a1f8c6f5422c88eb4ca7f6b198e76104962918e59f7c558f,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,20.206186,20.2,exact,canonical-source,profile_c418354814777762766b7183,reported,unknown,,unknown,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,apex-agents,APEX-Agents,tool-use,9f6769abba3f7bbaabe5c61215d2ae11c719ca8d1a15c693de3e1c7c0e312563,97a6785fd637d46b40f6d69f4e3deb6fc82d635bf7fe000eccbfca90c8d12d75,c9debe73032802fb9ce488d44f09b424ff97a76151b7b4146ddb59958b243b4a,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,37.6,37.6,exact,canonical-source,profile_099c0ba8b21819b771db4fc6,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,vending-bench-2,Vending-Bench 2,multi-turn-agent,007f0c3c496243838690df069bba27b3f0d86bce3476da3cf371a3f8b420038b,ad5d2a7d52a458582795f391a51c306be68f7fc7031ef18f1a16edbbad2d60db,38be7dd9ac77751164079289b317909d6a2c57075ba33d292d0899260c6fefb5,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,5082.936,5082.94,exact,canonical-source,profile_2b04515110e6907c541185f5,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,2a0670a4b7647129092df5e97574030145ae375a8ff58450d7fb545337eeb701,438b158a4d68d87565dde9b17a3db3ff375316972f994aa0618349399cabac64,2ffce7450eb8681bc24ef818920ce220a7bc6b670e86ea76c61e0feced6cf104,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1046.01,1046,exact,canonical-source,profile_c418354814777762766b7183,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
moonshot/kimi-k2-7-code,Kimi K2.7 Code,Moonshot AI,open,92,weirdml,WeirdML,coding,4e759bf4605d0ff61586da645cfdfaf18a48c55f9a71012bf2d871a1f578b63b,7272f4f9ff1f3b6aa3e412336686d918e8b8bbc3498b6943396b9bb3c12fdf00,e3029412c8500dac68ebb1de4e59896852700e290bd965b86f9f8cc7d11f9fb9,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,54.120000000000005,54.1,exact,canonical-source,profile_85f25685b1561fc4451541c5,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,epoch-eci,Epoch Capabilities Index,capability-index,b43b8764509c472897d80ea73a2e6514a32522c39d707a05b7fc5d999a6ef66b,d958d152708c615b7fb6ebf3fb6dac695367ba2c91c98fe2ff4678cdd8398388,ac3a582014ccb9e1add6048f494bee24f454ad2985c7137efcf44a6956c7374e,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,146.85,146.85,exact,canonical-source,profile_90e56fd70dd0364120d5d708,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,arena-text,Arena Text,preference,35cf8aaad2b9fa2c59c8c8ec777ad9f0eff4c6aaf9764fa0763d41d4593426b1,1fd4b540020ed46a12db1de04ae6a6db310679cbd6db38e44fb711def8f409a5,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1441.403236489422,1441,exact,canonical-source,profile_062b1dd6a5a62c713b280490,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,3e8d2522565e34e4fb8783d591b72fcb487cbf8d0a9623b5efb70d6cb9c7bd21,bcd5d7cdee647bb9f08d1813095c2bab2574d416311f8e80d993c7869682c8ea,868bb776b8522904ff0c5ee9dd97bd51db6e2af1f9b16da3811eb93464730240,release_2f3c6e21d0f3616fdb78f627,HLE · Qwen vendor run,full-multimodal-cot-no-tools,accuracy,percent,true,28.7,28.7,other-supplemental,supplemental-source,profile_28d4327ed17739786e2e7ac4,reported,unknown,,unknown,,false,2026-02-16,stale,258,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-397b-a17b-card,Qwen3.5-397B-A17B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.5-397B-A17B
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,ae372169a0bef9c68f25ba26053302a2e6de87eff8a980936985dc8d84e2e9aa,794b9d3a14e9e32cfb945c0e62dbaabdf5120ecff4f9c1148c358b422e7b0182,f316e6da39745a3aeef81c5887a1dbc43344c449a6fe5c2ae3ab631f773424ae,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,28.962002,29.0,exact,canonical-source,profile_d177f1d7065fe8f8ce6ad0c2,adaptive,adaptive,true,adaptive,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,gpqa-diamond,GPQA Diamond,science,b77ec050f53aa28119260c972f657bfcc9921703abe51bea5508bebebe46e814,70ed449f60ac1edce0d07c38bef94c79a6c0c7d19a29491204d5dd7ed04887df,bd9de00bd5433c347bd7f68e4aa17f4dfa14742cda8f4713d0d2ebc0be75c1fa,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,89.292929,89.3,other-supplemental,supplemental-source,profile_d177f1d7065fe8f8ce6ad0c2,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,mmlu-pro,MMLU-Pro,knowledge,ce207ba88dc13c7a99aa5a3d58b8d984d8e570abeabe617e299acfb27c5b711c,e3570f95190160986b5ad0e26740a4eac24c380aefbdd2a5ed46b3f4657f928f,44f56864a490bb7582461036c068400061930b80aee3b634127a43235a3bd532,release_95dc04609d027072fa229d3b,MMLU-Pro · Qwen vendor run,vendor-card-protocol-undisclosed,accuracy,percent,true,87.8,87.8,other-supplemental,supplemental-source,profile_54abff30262a823e8f5a7fc3,reported,unknown,,unknown,,false,2026-04-22,stale,193,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-27b-card,Qwen3.6-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-27B
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,9a7bb916d010611e847a7a8815fc725df23da7d3e0224ad158430431d04337ff,d45c1c519352dcbf0ba91d67ea73b35fbd636ddd63788d84615072db43d26a4d,f767e31ef1f628d16486a10f8c816a461a1bf1fc45e3fd6d62bc0494285aef5a,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,88.88888888888889,88.9,exact,canonical-source,profile_7cce63965b2039e8ed31c47c,reported,unknown,,unknown,,false,2026-08-07T00:12:18.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,frontiermath,FrontierMath T1–3,mathematics,4a4589d6b746837d91f663ab89ad5e31230d46634f8cf0e4c27274f9e4a35267,e2ff315674937383987073fd3aa42e03f1b6d889045cb33bbbb0070b4b96af36,d8ecfaeb0fafa1b6bb87324c051d31193978f4f8bbf35f3aa4df329071cebd42,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,31.2280701754386,31.2,exact,canonical-source,profile_f05cee9da83990f5cc4a1d0c,off,off,false,none,,false,2026-08-30T22:34:18.000Z,fresh,62,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,mmmu-pro,MMMU-Pro,multimodal,c32eb82db435f498be90eb9b34703e479609b080a33624d7fdf63bc823cf893f,d1f784aa5e6d21cc56b3dcd56661c807653e54eac23a93bf5f4e450ebb8765d9,f6a42b3c673b70f19114b664286aa3515baee4fb0558d019697ada3e4ef1f1f0,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,77.283237,77.3,other-supplemental,supplemental-source,profile_d177f1d7065fe8f8ce6ad0c2,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,critpt,CritPt,science,19f28c0d04bd7952bf8cfe0c748ed486580c40f5499cd83e55bacc8340a63499,b36880442937c4bb62dc47b7e369d64a043559f886c841522dfb9a652513e292,9eea1925b989dc40c0ec9008c754919603aa0721eaffe8b990fe19d8966dbaa8,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,0.857142857142857,0.9,exact,canonical-source,profile_f05cee9da83990f5cc4a1d0c,off,off,false,none,,false,2026-10-28T21:25:03Z,fresh,3,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,swe-bench-verified,SWE-bench Verified,coding-agent,96c3d27aadd523c2bbdb530cd5da43a329116dd83a0c30a0a97d2861c1c2c925,c2f707b99a2815e07c1030e5d1d1d5b3c12a1582bccd8136cfdf4c6caed5432f,76ed278b99ffd1e5f9230954bc4ee5268b74c5c42a740c706366f11a530ecbf6,release_9162beb27813d52a46a20a9f,SWE-bench Verified · Qwen vendor protocol,qwen-internal-agent-run,resolved_rate,percent,true,76.2,76.2,other-supplemental,supplemental-source,profile_54abff30262a823e8f5a7fc3,reported,unknown,,unknown,,false,2026-04-22,stale,193,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-27b-card,Qwen3.6-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-27B
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,swe-rebench-v2,SWE-rebench v2,coding-agent,b359c40930a830e9d1647d42947ed8fd8e385004370531197e9b0b3458cf1a15,d5f0bf44ff1ea49a9ecdca3047d6021cb335d747e2dd7a384178604ea738d8f3,01da0eb2fd9c383f12ef300cd69d69d04eda6900ff1aa1003fe2c725832a2382,release_7449132b5f078c71f23c26b5,rolling-v2,2026-02-15/2026-03-01,resolved_rate,percent,true,56.837606837606835,56.84,other-supplemental,supplemental-prior-window,profile_ddb434ba4c54ecfb0e035efd,reported,unknown,,unknown,,false,2026-03-01T00:00:00Z,stale,245,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,scicode,SciCode,coding,1504e6e001f80ef3d8b9fbae91fc307b7b1e50e1b937fa6012d8531f0d6ef7d8,581f164c27bb71270e14b75312a9c05bcf71a704da99afa82e77803f1950b4f9,8f1633398f8a29186d3ec6e1e9998e315767bb63261e54a062fc855688c57796,release_de40c66bd40a7b5483b40f9a,SciCode v1.0.1 (Artificial Analysis),scientific-research-coding-subproblems,subproblem_accuracy,percent,true,44.791667,44.8,other-supplemental,supplemental-source,profile_d177f1d7065fe8f8ce6ad0c2,adaptive,adaptive,true,adaptive,,false,2026-09-08T00:34:38Z,fresh,54,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,webdev-arena,Text Arena · Coding,coding,aeded49bc462d814813d99e8f6b2ebaa634ca594c84fdc254fca6eb8c3c1dc78,166f023084fb6e7f28591d17bb9193ac6f7cfdbc57d428191321a47ce0cd0188,5ba244c43b5f7537b5fd92430355ef77d604108fb710c111057a9a770fdb33fe,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1400.2208309279806,1400,exact,canonical-source,profile_4837b3098d8faed90802985c,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,9fa46f1672726652ceb4f95bad745706cf3097ead8174ebba497049165649513,1f87088a4d9990a8d2d18d331038a634a295ebcc4dcd2ca35d8f92c55071592e,00b43fd9ca289c42d0beb098afccdf3cf238f10120d2858968ec3d98c6b09f11,release_c53c830678318dc1763793c8,Terminal-Bench 2.1 (Artificial Analysis),terminus-2-89-tasks,pass_at_1,percent,true,51.310861,51.3,other-supplemental,supplemental-source,profile_d177f1d7065fe8f8ce6ad0c2,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,799e892b8f2bd742ec90e9f0d2d308a89b9189e793606a42b4e2b745eacc7785,3deb4d62d9413a7fb937cb2c33e270d4d8688c817cc9e22bf9c356d7426ca7bb,8f78486beeddbb49c8623a4314b10df5c24398a38651efea42424aa8802acc57,release_b7d9ab744ffac0cb856e114d,Terminal-Bench 4.0 (Artificial Analysis),mini-swe-agent-66-tasks-500-steps,pass_at_1,percent,true,0.0,0.0,other-supplemental,supplemental-source,profile_d177f1d7065fe8f8ce6ad0c2,adaptive,adaptive,true,adaptive,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,cfaa016efce5b4749862a903bcd48a3392870f87a952d62078ffb3c0e7e3c079,f43354c00908a8771d9b98a7f9e7240659ebd888cc8b6c83546bd5dde1199b7a,d6903f51b87e19042dac19fbb1d86f91bbc9d8ad83806b456ecb0ebe95e16dc1,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,13.402062,13.4,exact,canonical-source,profile_d177f1d7065fe8f8ce6ad0c2,adaptive,adaptive,true,adaptive,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,skillsbench,SkillsBench v1.1,tool-use,bdfed38e704e0c0b9349070d11f6a7a744e5b63f20f7e377d82e0cd62ae74244,f4fe44d781b8822dc371f22076a32f5ac304b518eea29f0ca81e0e11c7677aca,0075a32b62b602932d1572b58a2560fe97ac92d58166e3cd33e67ab077a6bac5,release_5af9fe5a9d5e275323798c8b,SkillsBench · Qwen 78-task self-contained subset,self-contained-78-no-api-subset,pass_rate,percent,true,30.0,30.0,other-supplemental,supplemental-source,profile_54abff30262a823e8f5a7fc3,reported,unknown,,unknown,,false,2026-04-22,stale,193,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-6-27b-card,Qwen3.6-27B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.6-27B
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,apex-agents,APEX-Agents,tool-use,efc3d2e4a1db10432044885d5c0d15d1e562fa5ca77b1e5ba1e5fc32c1a4cc78,1224deaeb2e1bc3bc83b53213416dc1dd94ccb205910df1d8496491222b0f2ba,8e4b2b2b9208e63ae9b3331b08fb7e1e68674f25cd35a332f6dddf35c0e78f17,release_e102ff7c1fdcabae1b00cd4a,APEX-Agents · Epoch evaluation since 2026-09-30,480-professional-tasks-33-worlds,pass_at_1,percent,true,24.9,24.9,exact,canonical-source,profile_f2df9bdc5ed37041104d582e,reported,unknown,,unknown,,false,2026-09-30T18:53:14Z,fresh,31,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-apex-agents,Epoch AI mirror · APEX-Agents,https://epoch.ai/benchmarks/apex-agents
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,bfcl-v4,Berkeley Function Calling Leaderboard V4,tool-use,83dd3b78c54352e64b8bff8bd90b81c57ef3bfc1d9658f94cec1b90bebdd251f,d7cb6a43287918f300482b7aabb4124578f3dff767b00503545c975d2b0a610d,2be3c8fde04bf33c6b6db13a6ae1981f288300b9baf0dde2bfae161e55ae4181,release_6b9456463dae72f95fadf6b3,BFCL V4 · Qwen vendor run,vendor-card-protocol-undisclosed,overall_accuracy,percent,true,72.9,72.9,other-supplemental,supplemental-source,profile_28d4327ed17739786e2e7ac4,reported,unknown,,unknown,,false,2026-02-16,stale,258,2,B,2,B,2,B,official-model-vendor-report,4,official-model-vendor-report,qwen3-5-397b-a17b-card,Qwen3.5-397B-A17B official revision-pinned model card,https://huggingface.co/Qwen/Qwen3.5-397B-A17B
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,23036c6d817b1044e889941e5a65ffff15331896921783abe3fa869da492f690,b3fd68385df92d6cb937b3010e8236767bd58a70f0d2670d2ae01c3d4add41ce,ea714612113691365b594e9602c7fed643989edbab7ab049f89a82d36993198f,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,800.64,801,exact,canonical-source,profile_d177f1d7065fe8f8ce6ad0c2,adaptive,adaptive,true,adaptive,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
alibaba/qwen3-5-397b-a17b,Qwen 3.5 397B-A17B,Alibaba,open,93,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,8b355f074dccabde9ceeb5a4bb26bfccc17a78678a244f55ac7e55afa5186e11,aa004272b65d364615cf5f98fd0b53f9dc824b734a3fa36169c50ced77388944,f060008efbd1330fcb16847cd355ebaffc070cbb0b9dd1b31485f1394c535163,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,9.79,9.8,exact,canonical-source,profile_fec2e8c4ae6e264376e8e9fd,enabled,unknown,,unknown,,false,2026-02-27,stale,247,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
alibaba/qwen3-8-2-4t-a95b,Qwen 3.8 2.4T-A95B,Alibaba,open,94,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,42b915cdbbbbb3884ad5a8601c0df9828227b19da0f0f200ed1cad13dba273d4,8ca7498bf7d5f957fad20dfaefdfedbd5bdf33488010adc22da9fca788313845,e1f956386a97e6ba42a3c396757902d19f19cdabb10fcc91e345af5c771a2890,release_fa9e908ac04b067eb00182d4,HLE official leaderboard,published-leaderboard,accuracy,percent,true,43.6,43.6,other-supplemental,supplemental-source,profile_9a07d223bdd9e8b87c6a58c3,reported,unknown,,unknown,,false,2026-08-29T23:52:43Z,fresh,63,1,A,1,A,5,C,self-reported,6,self-reported,hle,Humanity's Last Exam,https://huggingface.co/api/datasets/cais/hle/leaderboard?limit=200
alibaba/qwen3-8-2-4t-a95b,Qwen 3.8 2.4T-A95B,Alibaba,open,94,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,1ec8c8eed8f554bfdabb60ea538b86629f8000b50dc146313f55dc9d70810ec0,f7412f0a089a724590586d6a0328bed82fa58dbd510541ba6af8b27a5fec9bc2,2702d49cca72c3306f1d671402ff6b60bc98b2efd30c9305500933280881dfef,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,42.44671,42.4,exact,canonical-source,profile_dd2002eab3ec3258a39b1020,reported,unknown,,unknown,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-2-4t-a95b,Qwen 3.8 2.4T-A95B,Alibaba,open,94,gpqa-diamond,GPQA Diamond,science,895065af9442feffed6fdb928bd0f11d8823f949fa5bcf1ce9dcb4725286f228,61511f694177c3ac5442163433e507233122b620c2bd4d0b5e866ea00cfebf79,c951e008fd1e95da37f9853a35a9d63c98520d3c85b7c74a6bbef9ec347a8142,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,93.535354,93.5,other-supplemental,supplemental-source,profile_dd2002eab3ec3258a39b1020,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-2-4t-a95b,Qwen 3.8 2.4T-A95B,Alibaba,open,94,critpt,CritPt,science,16a4cabcd11c005ca4ef485820a322d57216dd0779b7328bd8a604a7c61820c9,330cad8965b17dc572a6d2256f66452d5f7d5237be0f260e5d2af44611ff03a7,32c64e7085a4b74a03105464f87aeab6550cfc6c70d84c111a6042eaa5aadce7,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,20.0,20.0,exact,canonical-source,profile_dd2002eab3ec3258a39b1020,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-2-4t-a95b,Qwen 3.8 2.4T-A95B,Alibaba,open,94,scicode,SciCode,coding,fcd0b9bfa39b5093ebf9065f2dc8ce9241989e4cbad3988c76deefa80e4a9e2b,57379139d5c1e5be64f81bf7c27532f0c6c7a9b2762d5a867561f2cf2f9275ab,a845e32771dc4669b5899b30646a47a4ffd982b194519955ab3c2f609eaa32dc,release_de40c66bd40a7b5483b40f9a,SciCode v1.0.1 (Artificial Analysis),scientific-research-coding-subproblems,subproblem_accuracy,percent,true,54.050926,54.1,other-supplemental,supplemental-source,profile_dd2002eab3ec3258a39b1020,reported,unknown,,unknown,,false,2026-09-05T17:21:27Z,fresh,56,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-2-4t-a95b,Qwen 3.8 2.4T-A95B,Alibaba,open,94,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,c3dc7dfcd629efb8bd0ea8274bddb504c527e7cff145fa81db1b4869bf7a6259,bf975bfef830870c8130cc37b217f517fef8833005b94ad809b4347582d8cf55,7bfbe2612c758d5760a631cc6ec05da1c9a0e2d8fb6afe9ec1f541bb80c98c4d,release_c53c830678318dc1763793c8,Terminal-Bench 2.1 (Artificial Analysis),terminus-2-89-tasks,pass_at_1,percent,true,82.022472,82.0,other-supplemental,supplemental-source,profile_dd2002eab3ec3258a39b1020,reported,unknown,,unknown,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-2-4t-a95b,Qwen 3.8 2.4T-A95B,Alibaba,open,94,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,c2602a60ecef8851a5f10fb933893989c9245bc0f82bbf0b76041914b808cc83,7dc4b699c46ccf03baa0d7670002f41c739c7bcffd38c890e759dfe794cd20df,78742f435bcf919ac96006668fc210d2024c724feb5db534903d520fdf88d8c3,release_b7d9ab744ffac0cb856e114d,Terminal-Bench 4.0 (Artificial Analysis),mini-swe-agent-66-tasks-500-steps,pass_at_1,percent,true,11.111111,11.1,other-supplemental,supplemental-source,profile_dd2002eab3ec3258a39b1020,reported,unknown,,unknown,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-2-4t-a95b,Qwen 3.8 2.4T-A95B,Alibaba,open,94,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,ff31f5e5f2f5dc63a328563b9e0743f81c46c269de987467397175b83756057d,5d2995595c233b13635135b080e23b009eac4d20ad9b54ef50f869eb064a4cb4,8990bbfc9edee19a0378f5574b8299d090877c08324a7f8eba51f2d76af190c8,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,49.072165,49.1,exact,canonical-source,profile_dd2002eab3ec3258a39b1020,reported,unknown,,unknown,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
alibaba/qwen3-8-2-4t-a95b,Qwen 3.8 2.4T-A95B,Alibaba,open,94,mcp-atlas,MCP-Atlas · Scale,tool-use,9ec65ee370d2a41eddc6b99d6586694edbbf2e4481919bbc84b82114d9e67948,779aab3ebb1f4a7a0b1aa6a062dff47673e0f4d58736e12dbdb1d83c62573844,e0d14641e671e7bf4e1c80956229c3a642815f6c6b3729f48d5be5d90baa61fd,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,84.5,84.5,exact,canonical-source,profile_99463af9d053c64e40c26c38,xhigh,xhigh,true,fixed_effort,,false,2026-09-17T14:59:32Z,fresh,44,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
alibaba/qwen3-8-2-4t-a95b,Qwen 3.8 2.4T-A95B,Alibaba,open,94,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,802229938acdc632d6ffed545bec486a7af4e01d79133c11b83ffc56f97343a4,1a9fe59010535609f0e640c1ca28be6a47fd7454822a46b9be34090728f73979,e7a3be2fad89121cb026b4893bb3c0d272e694a63588ef23570a75e241c538b6,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,1615.72,1616,exact,canonical-source,profile_dd2002eab3ec3258a39b1020,reported,unknown,,unknown,,false,2026-10-30T09:17:57Z,fresh,1,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
alibaba/qwen3-6-max-preview,Qwen 3.6 Max Preview,Alibaba,closed,95,epoch-eci,Epoch Capabilities Index,capability-index,1c79ba4f8cbb19572c5ae515f932eaf47e69770d7b8f0b3f58671394a4e434e2,bceca9151cad54df18457fa3e784af43f451b2f608f329d918c767901d115636,873b1028434ebba2742a925a924197c3ee9a3603d93041f13ab4bbf823508efd,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,149.38,149.38,exact,canonical-source,profile_d3b3c1e009228ac3e1f855f2,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
alibaba/qwen3-6-max-preview,Qwen 3.6 Max Preview,Alibaba,closed,95,arena-text,Arena Text,preference,c9c19a5b175e96a82e791801e99d5ee9ab3d05451421a8c30b1ca86769b418b6,1ec20d56906b3e84ad975527995e86aed4e91542191b433281a66ce26a4721ff,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1460.2131590348529,1460,exact,canonical-source,profile_e0cdc415c5b29e9a4e97c44b,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-6-max-preview,Qwen 3.6 Max Preview,Alibaba,closed,95,simpleqa-verified,SimpleQA Verified,factuality,158c92862098523075c051dcdbf5a23afa5bc5932bed8f6318a58e6989f8e77d,e1633df1ffbe058fc72ef08e56c918e41c722b52d2427ed1e27e107bb49ebc17,d333fa9c9c40094adf12bb365e2ca5af71c35f022bcf022cf3ed1e8461311ff1,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,52.00400801603207,52.0,exact,canonical-source,profile_3fc6875ee3cff2a0aacf3f9d,reported,unknown,,unknown,,false,2026-08-27T19:33:58.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
alibaba/qwen3-6-max-preview,Qwen 3.6 Max Preview,Alibaba,closed,95,gpqa-diamond,GPQA Diamond,science,23cbbfd21dee5765870c0f9ea0b54c653f6e947c62b56bdf911e2a17320bd11d,fd2c603d90fc2608fbb39cb254591004168833a494af0f23d78fe4b306e4c9bd,0b0b7941b04d488dd65911e62a5ef48434dbfce3fef6c7a28590d294277391d3,release_130743b4d612a7ac7c996be4,Epoch independent Inspect runs,diamond-epoch-inspect,accuracy,percent,true,87.37373737373737,87.4,other-supplemental,supplemental-source,profile_c656be424b91b5bbd75f2aec,reported,unknown,,unknown,,false,2026-08-07T00:46:39.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-gpqa-diamond,Epoch AI GPQA Diamond,https://epoch.ai/benchmarks/gpqa-diamond
alibaba/qwen3-6-max-preview,Qwen 3.6 Max Preview,Alibaba,closed,95,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,03ba39e15b251678319319ec9ed0026b43a353573e6cc9734eff30cb0c0a2984,e7026cb4d98e9157aa78c42276885e8e7bce89d3ee9b556b33e2866ffb8a8094,53721f0f1d9a2b4915003b28b41904796c4720cefb18cc3ec8c3c5237db1499a,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,91.11111111111111,91.1,exact,canonical-source,profile_4b807b8f467caef3423aa98d,reported,unknown,,unknown,,false,2026-08-07T00:42:24.000Z,fresh,86,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
alibaba/qwen3-6-max-preview,Qwen 3.6 Max Preview,Alibaba,closed,95,simplebench,SimpleBench,reasoning,49cc42c853febbbe9ba9cc88db6dfd5b8c517b58cf2e2e61d67e6a753812e7e8,fbd2d6f82325fae5f980ded8de6d3c531dcf1e4609aab1711c36ef05b0d279c2,6dbccd280fb64a450c00c0ff2232ac25f842fdf586a8d550543ba0310005b1a2,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,63.0,63.0,exact,canonical-source,profile_f0189e6e512a7f8f6c14d55a,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
alibaba/qwen3-6-max-preview,Qwen 3.6 Max Preview,Alibaba,closed,95,swe-bench-verified,SWE-bench Verified,coding-agent,4f7a2e18825248e23d2e9b9af78f71171c73247ff2c1681b0cae593c824c762e,6ffb16487a50a714ff3c7c929d863922eb54ab6ba9d38ed9c54ea6fd415f4af0,3fb6eb7fdcc7d7ef8a2ab537b37bdb23ab31a934113d4446ccc777d5c46c3238,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,72.8,72.8,exact,canonical-source,profile_1d20a31e24960743cd851fca,reported,unknown,,unknown,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
alibaba/qwen3-6-max-preview,Qwen 3.6 Max Preview,Alibaba,closed,95,webdev-arena,Text Arena · Coding,coding,dcc700ba0a09e3af1f3453082859a39277d7a7b8e11436a0133715593a3807a0,442b790298f1313876668934ca8c02287ecdb34e07c5e3bb95d54849baa0a521,89a29151fdb7e16187487835798dd5eb7f6abfa6ee84a0b58e7a608a4a20be31,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1481.8179582284795,1482,exact,canonical-source,profile_a0689db8baf849f97eac474b,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
alibaba/qwen3-6-max-preview,Qwen 3.6 Max Preview,Alibaba,closed,95,vending-bench-2,Vending-Bench 2,multi-turn-agent,7d3df7776d36216d79b3f345979373292d0e57745233d02398b87c76200bb095,ff0c90291dbc88e03b37028e194fcef5f386d6b6d951645c55a87c1d6f3b0671,ed54e35bb086ed6ae1376be374c47344eb3021e2ebf1fbbe3d1e4469cf6bafed,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,4254.1866666666665,4254.19,exact,canonical-source,profile_a08522d91e6bb53b70645ffa,max,max,true,fixed_effort,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
zai/glm-5,GLM 5,Z.ai,open,96,epoch-eci,Epoch Capabilities Index,capability-index,231d21570010e24d28e297738164d5959d3f1cf2b63fd351e9d3bbf27f788575,91c7c96cceab9b93ac1ed5d9ece19c52c8a933cf464a5748764b2316fdd2108c,9760fb7bae977280edb83ad403592430cc08dde477d094c98a90c7e2bd228b3f,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,145.79,145.79,exact,canonical-source,profile_85a8ec980f1ad054744d25c6,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
zai/glm-5,GLM 5,Z.ai,open,96,arena-text,Arena Text,preference,e6d79220af87f38316ce4993937d2c433236b6706f4e8a46b3f55c39a76974c6,86e93a76da0884a3309b1584eada007310fb0cb97bc017af17a94e9761f09b77,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1457.2370395963933,1457,exact,canonical-source,profile_0f408aa2a737e269f251bd52,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
zai/glm-5,GLM 5,Z.ai,open,96,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,31d766a61c06222f617f27f45763c8c99d237161d5bf21e26bf3118a45a69eaf,8757c3359b588d8e3fa27fdc5b2a6a5e79ad0df932ae24be959ea4d9d548b96a,e1f956386a97e6ba42a3c396757902d19f19cdabb10fcc91e345af5c771a2890,release_fa9e908ac04b067eb00182d4,HLE official leaderboard,published-leaderboard,accuracy,percent,true,30.5,30.5,other-supplemental,supplemental-source,profile_8bc699530339dcbdd472b963,reported,unknown,,unknown,,false,2026-02-13,stale,261,1,A,1,A,5,C,self-reported,0,self-reported,hle,Humanity's Last Exam,https://huggingface.co/api/datasets/cais/hle/leaderboard?limit=200
zai/glm-5,GLM 5,Z.ai,open,96,gpqa-diamond,GPQA Diamond,science,72a9cb01082e0fd66f3c54e8110738bb296de676f5c898feb6876b06d6112b3b,a80e977f69d13a35fa79988dc11a6577901e145a8a2b037bb83e864aca6b785d,84bc7170d2cd2a10bf4063f3368c91dd9726ae1a51b414f3c065e088a9005833,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,83.333,83.33,exact,canonical-source,profile_c3d225c30664ec01fcf514bd,adaptive,adaptive,true,adaptive,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
zai/glm-5,GLM 5,Z.ai,open,96,mmlu-pro,MMLU-Pro,knowledge,01c280936861d78a3e8b9a5f26e744fe12a7892aa54ccf16935765ef850ce778,95ff32f07bf828895cd409262ad99985ce047610f3d7ddebd3f55ca708a87a9e,7255300ff750a1be144dd15aaf255edcd95b8de423a1eb180d93c718fe5ba721,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.03,86.0,exact,canonical-source,profile_c3d225c30664ec01fcf514bd,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
zai/glm-5,GLM 5,Z.ai,open,96,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,4d44ef488103f8e5036a165845d63c21677fbec90fe93aafcb3620d48a370c25,af2d5632db64c19b400f78e46c1792f8cbd8f9ea4981c8b20bb760ec17174d6e,24834bb9e849cfb43ac09a1b9c8e00f0dfb2ce51bb07c739eb6f0dcde7817da6,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,80.0,80.0,exact,canonical-source,profile_47f623c312e058af8554fde4,reported,unknown,,unknown,,false,2026-02-12T16:08:22.578Z,stale,261,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
zai/glm-5,GLM 5,Z.ai,open,96,arc-agi-2,ARC-AGI-2,abstraction,a5f00c260d83bfe46d6cb5b19f4c03a91cacedcdef26392ab55e44a0aaf6c4c0,d8a8d1f6ffa3a5f40bb48b65d58d9c22f84bfcb4bb6c58bac2ecd10baac0008c,a5ac13d6ba1560e9e9fe4f99107a0952e138420cd6e6f9258afc35b2f41425cf,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,4.859999999999999,5,exact,canonical-source,profile_476fbefcb74061bff19d00f4,reported,unknown,,unknown,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
zai/glm-5,GLM 5,Z.ai,open,96,simplebench,SimpleBench,reasoning,82ab2344b8802e3f31520111661afbf9bd10a5bd9e457d0c268ab8295f1c1663,3e9157f5e61fa7bf97e00c3c0bd45331c88d9ef47a5d34aad9efe800825b9a7b,ba21d91c69b2bfb5562da24595fe712c6aa16a43bea96d18b26d0459c948a095,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,53.2,53.2,exact,canonical-source,profile_927bf51b0edb096333435e1c,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
zai/glm-5,GLM 5,Z.ai,open,96,arc-agi-1,ARC-AGI-1,abstraction,7d5534bae527b9c0bf40d2287f5bc8299c7fbc84ab5732f9b2c9a75305ae7978,c9c135b7c7708a93e4f6bd69bfc26b08e528026be7d67597f13e74ff3dc2f5fa,54207793d4d27d8f38a115533712e157798c696fa5d5368797e5c8a606bbf29a,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,44.67,44.7,exact,canonical-source,profile_c6d6445cb7c06336b725fc29,reported,unknown,,unknown,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
zai/glm-5,GLM 5,Z.ai,open,96,swe-bench-verified,SWE-bench Verified,coding-agent,efca3a916af92abec1f062e28449402df61683016645969bbb4851aa1572d17c,cfd4e693216e3d7df05bff959e88caae82588e85c6710899b132b639c91bb564,e5e1f04a928af175f9bb63792493818a8f4d5e94c6ffcdd00a598e6865f561d8,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,71.4,71.4,exact,canonical-source,profile_c3d225c30664ec01fcf514bd,adaptive,adaptive,true,adaptive,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
zai/glm-5,GLM 5,Z.ai,open,96,swe-rebench-v2,SWE-rebench v2,coding-agent,39af4b113159dd99644b030d734259a14ba4ec78dbafef15442807f608da7a95,7d6cdf91cf00098d20807b8d5fd993c19ce5780a669f1f5e519dc4c0adea7078,3a19811f492e0107b51dc9de4ab7d657eafc63c27b4c368713b29690d8fd2d14,release_7449132b5f078c71f23c26b5,rolling-v2,2026-02-15/2026-03-01,resolved_rate,percent,true,62.96296296296296,62.96,other-supplemental,supplemental-prior-window,profile_d4385bff71d0ba7da64cdc30,reported,unknown,,unknown,,false,2026-03-01T00:00:00Z,stale,245,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
zai/glm-5,GLM 5,Z.ai,open,96,livecodebench,LiveCodeBench,coding,2ca00aa602b75605dfa9d2f3f574962e397c56da4828973c5a12c826f4b39252,6cad06ef23fbe63ecfd74c932f34baf0335335656fee697acf36ae3773a4a587,d9665a74984bdd9123ef1ef4b855a575fb8ce7e6e334a426b44b6f9ecd38e722,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,81.868,81.9,exact,canonical-source,profile_c3d225c30664ec01fcf514bd,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
zai/glm-5,GLM 5,Z.ai,open,96,ale-bench,ALE-Bench,coding,71a63c3fa42f889333295fad7243a3fc7cd0131ea674c5edc5e50ab108f67e67,161cff83cf84783e5cf4644b5207e1e1e35376b7f65192fede1c11f8b9dca8ef,1e275173e9df9ea2c30f14c0e4badff4ff8640ece726084771e1affe20ef4804,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,765.62,765.62,exact,canonical-source,profile_a6a581cf3925bd5dfbd5710f,reported,unknown,,unknown,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
zai/glm-5,GLM 5,Z.ai,open,96,vibe-code-bench,Vibe Code Bench,coding-agent,888f2fbe69b06eedcd4ee90235b87212f2af83739c25120c77c4bf635bd81bad,8747de58678dfd5f9193c700196d1fcf5f9d6da1e40f76c3c509a1fae798294b,fbaf9d05cbc62c85cf56b0a4090fbe24f3ed44f21be79a7bddb66ef8597f7ade,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,23.359,23.4,exact,canonical-source,profile_c3d225c30664ec01fcf514bd,adaptive,adaptive,true,adaptive,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
zai/glm-5,GLM 5,Z.ai,open,96,webdev-arena,Text Arena · Coding,coding,e203b80d4475687a7f64cd50772b0e2e57151b069c2e8575d904811efee72595,d5042ff681871bd502ea77ee9159d1ec0dfb8a71589e55a79af6e4b95585d316,bba3a0a1773262918b49cea5234f83ecb793964f60b767df6d0e9a19c6bba4de,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1435.569438296295,1436,exact,canonical-source,profile_fcf17c516dbc641b9250907f,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
zai/glm-5,GLM 5,Z.ai,open,96,vending-bench-2,Vending-Bench 2,multi-turn-agent,95c5f50e4f834d538747477f06bc4ef2202cc452738e82363bb8fd82621cb83e,e6262b1eb3d735be11cc784679eb3728eec49e12f3b0bc10ee599c00af353148,554f9877bfc4e217d97cca49b51d650a61d333d955822e66a234c870093a445d,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,4432.122000000001,4432.12,exact,canonical-source,profile_cf429bf4453c02a3a83fad37,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
zai/glm-5,GLM 5,Z.ai,open,96,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,3ef7c6432ef6064baa419a68bd91a7ad41941afbc61ff08ce06213e51ef02a12,fb288f278955b3b9d506e1b53929591fbd647e671864b45a9bc55d17a1b2aef6,d56ddeb73571ba07e4be83089b0896e0f20a4d20ab1108fa0607d3706ea2a01c,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,9.79,9.8,exact,canonical-source,profile_ce13d8bd3f6566efc90ebe09,enabled,unknown,,unknown,,false,2026-02-27,stale,247,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
zai/glm-5,GLM 5,Z.ai,open,96,weirdml,WeirdML,coding,0274294a6c12935e0e768ca0d3c6ca7a4dc41ef7899119c31e7eae8216fc7c54,689bd3af35d1ef34de07a75f5794fbb623177429e5f075da17219a465fa2d3c2,4c2a0f437452a47939588543ed5de59f8baa840b89d7cdfba721482c4cfae020,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,48.17,48.2,exact,canonical-source,profile_818aa5ca47a72c9303b2a24e,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,epoch-eci,Epoch Capabilities Index,capability-index,2b5b1c8d482b43c12307931a09a37f99d3be75de5a816febddc369ed89e6e3f1,364941eabbbd085643141eb43df49f1a06f343edc30bdbf96484964a1c8aa252,59347700fac44206bb52fe80476be25aa9d8da91467d2e8dd0382d5ca6deb7e8,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,146.8,146.80,exact,canonical-source,profile_d4c6e9282aafa0f6c5e6104b,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,arena-text,Arena Text,preference,0a92ab3fde14edf41463536976e2737b16db3768cb3c4969f2f439066140d108,327a17f91a10e363cddbbfb564fb9000062e34d414791b4b57184c4674f7f0b5,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1456.3246398176152,1456,exact,canonical-source,profile_cf2150debee34b4d5f1bf4b9,high,high,true,token_budget,32000,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,simpleqa-verified,SimpleQA Verified,factuality,7212929dca6875e70f1d45e89b7f0b724df10e74185bacfa0e8b73c6dfb6e88d,5a8c6d572c6e46b19102287b076122ec781ec26e1dc13ed41e7b712e9dfc3881,17365a273d377061f9af1db8b6ceea460b5ef524be9d11495cc86c480c6dae13,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,23.7,23.7,exact,canonical-source,profile_1ce9cf4baebaf695f5b05969,reported,unknown,,unknown,,false,2026-08-10T22:42:10.000Z,fresh,82,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,gpqa-diamond,GPQA Diamond,science,9d3054c2c587c743ab50277961c2d4ba73a3371dd37cd7baad0b28032b504cf1,8dee653c1f35ff5eb7d2b91bbc230bed33d29b572849fcaba6dbc6f8d3891a59,2c6ee499dbdf07f14ad443b7e13b5062c65a6b8a12fb81e1d2da59655aa6c4ae,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,81.633,81.63,exact,canonical-source,profile_642e3a37222b1abf74918908,adaptive,adaptive,true,adaptive,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,mmlu-pro,MMLU-Pro,knowledge,3ffa7c291ec9fcc518a2f184f5b9299a9e229a6fb96c09413d06b4ce5bddc46a,a50f6fb9a9011342347ef4dccc9bf8566f0cba2c89b6c07da8bb48146b55dbe1,fe3e35cb3d9898f9395a5e010edd85176dd26026ac66a23654a2298ecbaa5a88,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,87.357,87.4,exact,canonical-source,profile_642e3a37222b1abf74918908,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,frontiermath,FrontierMath T1–3,mathematics,f11935c3b003fb2a4f6e004617bd1b082f3b7b10c842c1a17940a6e70dd35ed1,4a50bf3fd641df212b6a616f1e30a465db30ab659c5acc42521f53ff7379dfaa,2ef49575354bf730e05acc6ad0cd5064457f40531f1dbdd9ad25e49f1ceedef2,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,23.859649122807017,23.9,exact,canonical-source,profile_34919f24dc735cc9bea725dd,thinking-32k,unknown,,token_budget,32000,false,2026-06-11T21:10:25.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,arc-agi-2,ARC-AGI-2,abstraction,ee22e98cebc6a8f3bae398f447e25fc8473985153bd9ada379d8d59498c8ecc3,7cdf20953c18e5700e7f6418d51533caa1ee0ff811f555789aa820b757537c44,6b9c2fd7654c4f0f9bfb026fcfc72a0b96661576340141c91daef92c4047280f,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,13.61,14,exact,canonical-source,profile_beda1f27c08006aeb0f60b10,Thinking 32K,unknown,,token_budget,32000,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,frontiermath-tier-4,FrontierMath T4,mathematics,8329d1ded5feb70440816df5dcc2e63dc3cf204904f69672e339775f89c167b1,3ffef1983586cb775cfded7875b1c400a74aecacb099609f5a0cef878a3b026a,b012c400b53926d881db508924e56cc34a6750fad8cb9a2c32cefa8a1c6bebd3,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,2.4390243902439024,2.4,exact,canonical-source,profile_1749eab61af56fe6c62ab77a,thinking-32k,unknown,,token_budget,32000,false,2026-06-11T21:10:25.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,mmmu-pro,MMMU-Pro,multimodal,94eb6729ba56d1aa51b0b4a728a4608d868d618164d10486dadc051a357bf951,a6db2f818db5eccafc2802c3f413ec0e6c6bbde0602466cd3a9c78379fb9b0a1,acab67f0d779c48c32e1847f295c19d0ea2418192711a26094aa59ebd021a23a,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,79.306,79.3,exact,canonical-source,profile_642e3a37222b1abf74918908,adaptive,adaptive,true,adaptive,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,critpt,CritPt,science,bb3e72ef6a01474b23dbb3a421302af3988f00c2dda4ba67f6121a5a8b720922,a87022c58082620898f2d2e8786fc98054338a656ae805fcce087c680533a693,0dc06ea062155b9a4ba5cbf0d3ea98569889dd8b4f016dcf20079beca3db48cb,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,1.1428571428571401,1.1,exact,canonical-source,profile_3faedfeaf2b018b3d825930d,reported,unknown,,unknown,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,simplebench,SimpleBench,reasoning,ae2bd244bdb1584ac5da3d0ce1c236fa8115b99a8ec665b8fb7ee06a98826130,3142e05b214523871a6bd1a3e09cfe70b6cbb178067cb1301a9066c5b04a2ecc,5cd245fe69b46f91f733fb3a26f963748494b2b134260f93e466648553703f57,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,54.300000000000004,54.3,exact,canonical-source,profile_8c431ef1c3357f4933de067d,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,chess-puzzles,Chess Puzzles,games,2f162691456aee6435bc16e68a86d8a8dcde0fc116b87a0430868395e3486960,d4856e803f6701b17e6eb6b3b178339a4bc9361df2f02e66a76d59fe29343c80,9e10fa9584f4d877af324f2420f4db6e051b664a3d633dc8e22aeea815a71a6f,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,6.0,6.0,exact,canonical-source,profile_6a469e49a01c32309c33941b,reported,unknown,,unknown,,false,2026-10-13T13:43:54.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,arc-agi-1,ARC-AGI-1,abstraction,8da819e49d5586377af1d75b39c3ab1402192d9812946d88c0372360dedbbefd,60293b5967d2c3ece1adca411920a3ca1c26d511fe3b80023a1fb0846691612f,927a7a84c5b9eaaebb3c9c04a4e8184369edc2fa0aa7bae89f816e11c3f637c4,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,25.5,25.5,exact,canonical-source,profile_0de451a2a2c91a817f1dd3ec,reported,unknown,,unknown,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,swe-bench-verified,SWE-bench Verified,coding-agent,09906b3fd9593b9471c0a993ebab84d5d8a904eb95005f73d014018efceb2d79,5e3612fb5759c9b1136273503e9b2506bd47979da036a22618ef746da40af950,53806692743f1c24060f5200b18166415904dd2f1b3f27e3073b67829bc285e5,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,70.0,70.0,exact,canonical-source,profile_642e3a37222b1abf74918908,adaptive,adaptive,true,adaptive,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,swe-rebench-v2,SWE-rebench v2,coding-agent,fb2797a42649e5118de073994352126ab96b85c093ab401f0fee5d5e675abb9f,f80df3436e6ca59bfa5b94d001f53be83c97dba213069c91e7e77a983d9e66e5,afa1c5db8191e3968fbe9572001c1e06f3535024e3785db30a617ee2b341f91b,release_f0409f3e3e823115a9e78774,rolling-v2,2025-10-01/2026-03-01,resolved_rate,percent,true,54.0,54.00,other-supplemental,supplemental-prior-window,profile_8a4d978fd009422ee97e7041,reported,unknown,,unknown,,false,2026-03-01T00:00:00Z,stale,245,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,swe-bench-live,SWE-bench-Live · Lite,coding-agent,36cba32cffc68cf93eea67c914f50debcd4dce8e850ebecc1144a5918a862659,c65bd3ada82529bcd2795bbfca9a082dddebdabfe3904a0204fabb780c57c41f,d33b6160d8c4f5486c61d6719087c8fc1608c55248bf81a1934a392463fb972b,release_9d2de9d5be7e9822af19f7b3,SWE-bench-Live Lite,lite-300-python,resolved_rate,percent,true,12.0,12.0,exact,canonical-source,profile_4d43287ed026bd0acbdb0e15,reported,unknown,,unknown,,false,2026-06-29,fresh,125,1,A,1,A,0,A+,official-submission,3,official-submission,swe-bench-live,SWE-bench-Live · Lite,https://swe-bench-live.github.io/reports-0605.jsonl
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,gso,GSO,coding-agent,9f3898a313e2a8ea73acdcd9f468e11dc57826bd6be14876934f4e46eb2d44a3,61e31c5724f9630cd74a92bb4eff6a83c54caf3bc1cc49aee004e140a7d37a0b,d0523de259379b25230c65ac521c8505beb7b2d3c498b5f68d64e26dccb9991e,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,14.7,14.7,exact,canonical-source,profile_6590a3699935a47fe2f51b4c,reported,unknown,,unknown,,false,2026-09-02T09:27:04Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,livecodebench,LiveCodeBench,coding,5ebdb50f51a0678dc9402cdf99a643d154478236f5fd904cd135d164e5aa78b4,6cf4dd7321c68ab4be76822b95cfb7f6a736aa1b80f0eae715f6328d29b4697e,aa69fb327b16920f19fd28a36f2d74136082ac628e2d40ab9e92dfabcb3c67cb,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,72.996,73.0,exact,canonical-source,profile_642e3a37222b1abf74918908,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,scicode,SciCode,coding,fd5cd324ad7f434402b8cbad97a1f1de75e112466d82632556a4033771974d49,a51b216c47d6661d67e06f92044ce7bc48f329fc2d20d9a579fb68394b0773cf,17dab62fbce1564a83a9656d5ed96f4dab04756e23fd6b8fd62c27bb13d51030,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,45.717592592592595,45.7,exact,canonical-source,profile_6fe722919cfa837c93d9a4c3,reported,unknown,,unknown,,false,2026-10-14T07:48:12Z,fresh,17,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,vibe-code-bench,Vibe Code Bench,coding-agent,a0b5dcd6b67298f4bd74cffaab849221ada76f2a8cef45555d580351884b746c,b2356f238ca389c149d24431ac4e040639e7eaa5e382d9e435aac21afd03c3f0,7017681f55118948ad1826539c971c51957609d993b961c5955a8d66b14cc6f2,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,22.621,22.6,exact,canonical-source,profile_642e3a37222b1abf74918908,adaptive,adaptive,true,adaptive,,false,2026-10-08T22:23:23Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,webdev-arena,Text Arena · Coding,coding,3b82186e54a8332ab295ef4d5c6068fc9a6627d1c55ab26449ed6446f8a1498b,c9f0b41a61f16d9cdb650db728adfcaa4926a22a1092ecaf429fd72a0e70f317,3bfc93b849f4dd2f1976e05b3db4a8c9679101d1975e4377e856af36a85d3bcd,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1393.5404164128754,1394,exact,canonical-source,profile_cf2150debee34b4d5f1bf4b9,high,high,true,token_budget,32000,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,mcp-atlas,MCP-Atlas · Scale,tool-use,9c19239be878d3aad5df1c7844ba790860e590832a499d4193a4229c9b926d27,2e7fd8a59a087015176aa9468cab03354b7b86c58a67aacec092c75ef1ad5255,2618c486aa0fb0e79a31b09b04a577d89f45716f9bbb4e5b37e5f840073e52cf,release_6481c5033c8c5bc6dab70199,MCP-Atlas April 2026 rescored protocol,all-1000-public-500-private-500,pass_rate,percent,true,59.5,59.5,exact,canonical-source,profile_47645ffae78b747c6516878c,reported,unknown,,unknown,,false,2025-12-16T18:09:14Z,stale,319,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-mcp-atlas,Scale AI Labs · MCP-Atlas,https://labs.scale.com/leaderboard/mcp_atlas
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,bfcl-v4,Berkeley Function Calling Leaderboard V4,tool-use,8013768229a3cc785b42af94b8f2abd61336905c5bf7a9e08793b3dd2fe792b5,5138670fb30340f44ea041f2b385ac1d9423726f5b82fa67b2e7658de6b09e92,6a9c99bccbca11748bdf70a2bdd6eb864726c94bf4a029f84ab95898fe96d192,release_7d2560723dba9d549cd10b25,v4-ede5081a24bc,overall,overall_accuracy,percent,true,73.24,73.24,exact,canonical-source,profile_a5edc006937c8df2d2cc0448,reported,unknown,,unknown,,false,2026-08-28T22:03:18Z,fresh,64,1,A,1,A,0,A+,official-leaderboard,6,official-leaderboard,bfcl-v4,BFCL V4,https://gorilla.cs.berkeley.edu/leaderboard.html
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,vending-bench-2,Vending-Bench 2,multi-turn-agent,c1268e0d69edc2d740b6d73036d76fa28d19039e785028fa461d5ca8f5ce34da,71ebe1de50c6a5879a8d6174b987ad2428aabf5de3518be4a985144988badbd8,990637a02b64bd8eac62a2e1fc535612686efac216914c730ecfc273dbc4f6ad,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,3838.7400000000002,3838.74,exact,canonical-source,profile_f2188ef14be4e31531958a1b,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,98ba99e50cf883dbff741b0063aeff69cd361d8ecde0bc6e19eec0e68b39e103,49e883531384e48de606b1543f6e8ad30c1546a7e7d15bbafcbdc641a2e74eef,146ecd0614616970e59ca1f3b030d587e15097d46f410a08716a1dda9656be65,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,915.41,915,exact,canonical-source,profile_642e3a37222b1abf74918908,adaptive,adaptive,true,adaptive,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,gdpval,GDPval,professional-work,f6b38ea287a003bca656f4f20d6f595f4ec814e1a592ac6373c49421f1bdcba7,eb1ddddbe01d4883ee3575f694a340daaa0e6cf67b010e820664d30e8de76320,739dfefac2e9651bb5c3941cdd027d7dc9b413f02dc8f80441112582eb9505cb,release_730527fa346470703bd19e21,GDPval official leaderboard,expert-comparison-win-rate,win_rate,percent,true,42.5,42.5,exact,canonical-source,profile_2ef94f4849290d1cb68eb0ba,reported,unknown,,unknown,,false,2026-09-02T09:27:05Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-gdpval,Epoch AI mirror · GDPval,https://epoch.ai/benchmarks/gdpval
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,tau3-banking,τ³-Banking · Official harness,multi-turn-agent,6e03c270dd636117b551c23cc07598fd4e4652911cca75ab264bc32eb8e9fae5,8a773c0b52f5437bef7e0c4f4bab22e33138911aa1b6564948ca0979e706d35d,ff01b22ab20c5e42ea4cd21f5a09af629c356c4279c5f9921d42a8a43e64b968,release_dc4a6b19eb552354004439bf,1.0.1,banking_knowledge,pass_1,percent,true,25.26,25.3,exact,canonical-source,profile_4a73c29927d902c07689647e,enabled,unknown,,unknown,,false,2026-02-24,stale,250,1,A,1,A,0,A+,official-submission,0,official-submission,tau3-bench,τ³-Banking,https://raw.githubusercontent.com/sierra-research/tau2-bench/main/web/leaderboard/public/submissions/manifest.json
anthropic/claude-sonnet-4-5,Claude Sonnet 4.5,Anthropic,closed,97,weirdml,WeirdML,coding,6bbd151c4a04421ab456de1ecfaae3128940c94597cfa5274e19f4f04eb69e2c,fc83ea6eaf3e6c1a1444a5b219dac4f8527cea7e0f57530fb6949dd9495a2979,ec89876074787b4343e85caed455c416b2dc7bd739646f50439caa15ac87e753,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,46.7,46.7,exact,canonical-source,profile_ccf30264fde8350df3c208b8,reported,unknown,,unknown,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,epoch-eci,Epoch Capabilities Index,capability-index,323481449560a1bcd648b7b0c2059474c2495acd5bba0ee48d3d28195e99e99f,5316c8b927b9cdc2dfc134e4795722164785238e0c0d0c950679d6ad0286e535,a204e15366cc10c1300dac03f6867aa02125b761018aa4f31d9ddc2b0079cd80,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,149.89,149.89,exact,canonical-source,profile_feeea085ab7763bb52bc3cc1,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,arena-text,Arena Text,preference,e49492e214d5fa0427e3abd97c56928022f897a22b1a2f0ec95e0fd5ecf0db74,dbb76fa691d33c2b6198205a92f9fc854672443f25c69c1be0a535a2099dac67,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1455.7030125963045,1456,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,hle,Humanity's Last Exam · Full multimodal,expert-knowledge,74afab99ddd6c2af0cde19dfdec85cdc20c262a707039386407d757ae949567b,2d63369c94548977de9f234f730076427aa25c16420a0d25383cfcb22049638f,706a926b8aa5f4ce2838262333024c15eed3fff2541fbba585218c09e1cec2de,release_8817af9d9a0235bce0826e22,"HLE finalized 2,500 · Scale protocol 2025-04-03",full-multimodal-2500-no-tools,accuracy,percent,true,23.68,23.68,exact,canonical-source,profile_e92c503ca7d7c630bd73c098,reported,unknown,,unknown,,false,2025-11-26T20:24:19Z,stale,339,0,A+,0,A+,0,A+,official-benchmark-platform,3,official-benchmark-platform,scale-hle,Scale AI Labs · Humanity's Last Exam Full,https://labs.scale.com/leaderboard/humanitys_last_exam
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,simpleqa-verified,SimpleQA Verified,factuality,66ef421f73e537432bca3f44986bd95511bba32bf4619ac84d7d356826a145d3,c335dcf0dd43872eb06f5c4ddd5d319ce5a844a29dd65031235db8e61faa9f76,5ad4fb8ad8c4948f2431ce9816a32c2e10e39c3605456ba246ae3d4dc99f004d,release_7bf989f19fbe70ed22a4a50b,Epoch independent Inspect runs,verified-1000-epoch-inspect,accuracy,percent,true,48.0,48.0,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-08-27T19:28:24.000Z,fresh,65,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-simpleqa-verified,Epoch AI SimpleQA Verified,https://epoch.ai/benchmarks/simpleqa-verified
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,gpqa-diamond,GPQA Diamond,science,804708f17f8d8eb8f8febf0d2fbd79392b9857808583986b61c48930a930e757,81c799ff2e8136ae5981937a95c37df96d72ade6b3b3777e1d98fce3a58ef5ef,46906de3aae232bf19a1e0f1d696a6c4ed7f2312c92b10fe40be1db74ed195e4,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,86.616,86.62,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,mmlu-pro,MMLU-Pro,knowledge,de2540af20000fd9188239e78f7ed56be44eee2f2c37f06a56f0f0820505d65b,2d1178e8fdd84308fb320679e2a334043a4b6b3abb6bed7ecf9b01ca751e8a90,c598aacf77ca73dc50388c6275f2354c7cca6b36686847370e3441ecec78a574,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,86.377,86.4,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,otis-mock-aime,OTIS Mock AIME 2024–2025,mathematics,25f32caf886ecd5df67d252529348d2ae06e37575799439f8934b9ff994ea7b1,3c1ddf04203780ef8da5d2e6a94451a2af97d2e372256417a0c231dad923ace4,b2b66eb3c4c6788182a673f4582d4c4e64ce5ea5e7355e9721b1b7a059b65779,release_071fe253e79f87573f6d22b0,2024–2025,official-45-problem-set,mean_score,percent,true,88.61111111111111,88.6,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2025-11-13T19:52:15.595Z,stale,352,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-otis-mock-aime,Epoch AI · OTIS Mock AIME 2024–2025,https://epoch.ai/benchmarks/otis-mock-aime-2024-2025
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,arc-agi-2,ARC-AGI-2,abstraction,8262c9cfc169a68e3b698746caf974e04a1e7f8f0368c6fa7af6707ee9e72131,57a50c0d19780d4fcda3edaedabb572293761706401e12bdd96eb36084686463,e1fad15c71d90a914eccc3e146a77b63abce2bec16f2cf125a4be0372e9d72e1,release_1e2eafde2dbfd20577b0aed1,v2,semi-private-120,accuracy,percent,true,17.64,18,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-08-27T22:30:41Z,fresh,65,1,A,1,A,0,A+,official-verified-leaderboard,6,official-verified-leaderboard,arc-agi-2,ARC-AGI-2,https://arcprize.org/media/data/leaderboard/v2.json
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,mmmu-pro,MMMU-Pro,multimodal,b7bdc66f77a01d7e6a1b9da44f1075eab2980612ba5ae3e01f1fe3c572911e82,7be23ad7025b7368bde36ad1e9905de4959ed90c4f2bb6118b1d44a00538e339,44f96346dd2bd4fd48185f9c3b6d15fff4b677f23aa919d70d1d3aab0ed42b0b,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,83.179,83.2,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-09-02T09:40:41Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,critpt,CritPt,science,378a2a9f7260eb4fb0e6c0898e78ab5252779204789533624d380525179b9a59,14606e3cbfc292ef554859437b362d228284e9c022b89612d223194c834f45c3,0853f78b9e289cc0e250f444712e8dd3a8901363a4379256fd72468a457f8cae,release_079aaa34e0885e15655a2847,CritPt current leaderboard,71-challenges-checkpoint-accuracy,accuracy,percent,true,4.857142857142859,4.9,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-09-02T09:26:41Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-critpt,Epoch AI mirror · CritPt,https://epoch.ai/benchmarks/critpt
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,simplebench,SimpleBench,reasoning,3dfbb43cf68157a8244eb7e79f3e5c496aa11fadec681da6affc9a92c56f2838,4b16b3334fa01221f540df867cbc586722bf73c953dd1c4b07cf17d1a322e8c7,2ef4f664bc752406a303de62e789284339a6581f4537d42978aa376d823bcc66,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,53.2,53.2,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,chess-puzzles,Chess Puzzles,games,433e49fe271a4b36fe4879342a91a79eabd3c94919b546a3e8b1fe6c8e864f26,63af0784aa6e30312dfd994d8e0f55dac567cebbfffa468ee032be5f425d8108,431454578a2ed48fcee7e4c783986d5645a5b8ea55481c258c71960e1aebfc95,release_9f5f95bd857f9929b98d6528,Epoch AI Chess Puzzles v2,100-engine-generated-single-best-move-puzzles,mean_score,percent,true,42.0,42.0,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-10-13T14:05:29.000Z,fresh,18,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-chess-puzzles,Epoch AI · Chess Puzzles,https://epoch.ai/benchmarks/chess-puzzles
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,arc-agi-1,ARC-AGI-1,abstraction,f76d75f319838d97c8a4d7d69b3212b7fc629d361685c39dd6e174a9ca7110ef,4b18a6af4725f0f92e149f88e91e3dc344284b4d189d37647169e2aac6777dfe,eb04a4be031c049fb589d6777a50cf93f1483a73693dc20532c216d64482e27e,release_2e65e63d3852e53d984c2949,ARC-AGI-1 current leaderboard,semi-private-evaluation-pass-at-2,pass_at_2_accuracy,percent,true,72.83,72.8,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-09-02T09:26:47Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-arc-agi-1,Epoch AI mirror · ARC-AGI-1,https://epoch.ai/benchmarks/arc-agi
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,swe-bench-verified,SWE-bench Verified,coding-agent,1e1d682678f5810ac425368fa48599ce788aa0dd1e53f2da9f4b3fdd80d663e6,184f2c0059aa038dc3729bd7a73f2578a3e6f12c062a30d1aa58f98e39e77268,362b7ad838f32f83ba9345c2d0bc89e59f922a2565c2143c2e9cf194c867206f,release_2242ce390e0c2cd53f1b3a5b,SWE-bench Verified v1,verified-500-vals-mini-swe-agent,resolved_rate,percent,true,69.8,69.8,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-08-27T22:30:44Z,fresh,65,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-swe-bench-verified,Vals AI SWE-bench Verified,https://www.vals.ai/benchmarks/swebench
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,gso,GSO,coding-agent,c36ed8c687ac44142c5a7fdea3aca464866573846f2f460225c9bbfe36f91694,335601fd4d05cdebc6f3bf5b73fa70c4779482c0e216fdb8ecb9f5c18af8351a,01cb454390bd6d82cb9a851b14464ba4a4864e6d59cf0d74408c793b8ea05cf4,release_176f07ea4e9e0b8dfeede990,GSO current leaderboard,official-set-opt-at-1,opt_at_1,percent,true,13.700000000000001,13.7,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-09-02T09:27:04Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-gso,Epoch AI mirror · GSO,https://epoch.ai/benchmarks/gso
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,livecodebench,LiveCodeBench,coding,baadaf1f2639d51842977c6da657ada986e5160c7063766dd778d454c8e6fa52,42eb3e8a62e4e7918213f5c48bedc7defeda2f4166be00e71a9003502f9df62e,4195f0cb3428db1894426c72ad20b08f45ffcff313c8bf87dc83ee9259ea1ff6,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,86.486,86.5,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,scicode,SciCode,coding,d54fbe3f82015faf316c4ce7595093835df1880935e3047bffa2c4b2e3c161a5,e5dd38cfe41185fd2973437e9c46a5977bafd16e50abdd3e03bb7291f7a3dfeb,b1c2fa312286f3d10996ed85c71f12bd264c20e0b9ccbc43fe486fd6e266e88b,release_6d948ae7a2a0a695cbdf21f9,SciCode current leaderboard,scientific-research-coding-subproblems,subproblem_accuracy,percent,true,43.287037037037,43.3,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-08-28T22:47:54Z,fresh,64,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-scicode,Epoch AI mirror · SciCode,https://epoch.ai/benchmarks/scicode
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,ale-bench,ALE-Bench,coding,8a6e98ece7d8cd69519f91f59574ee5ff008d622c4ec8e2b7a8c207373c0afe5,aa96670f674d6a2c3b277e8fca43aec98a0c1c7505cb7011d6692d2edbeaf255,a6d89d9cdc6b737bffc5b0b25b7cef378a9abd72faf61acd5608974604fa4ad6,release_a4c08439e33767077f5fed9a,ALE-Bench current leaderboard,atcoder-heuristic-contest-set,performance_rating,points,true,1192.15,1192.15,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-09-02T09:27:01Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-ale-bench,Epoch AI mirror · ALE-Bench,https://epoch.ai/benchmarks/ale-bench
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,vibe-code-bench,Vibe Code Bench,coding-agent,21d6583342d88d7030ced45fd0326058260ec135f1b6438c4b2955ab1e0b234b,489c244c1ee48b39f9fa01abda4854cabbd7604e90ac7ad67ff120c0c802d8a4,e5bd652018221cf6cd3467f3047d879aeeaf3fdb89c340ec95e6c8bfd9b97456,release_ee527953b92b7b6f6f1beb2d,Vals Vibe Code Bench v1.1,vals-vibe-code-bench-overall,score,percent,true,24.606,24.6,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-09-02T09:47:44Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-vibe-code-bench,Vals AI · Vibe Code Bench,https://www.vals.ai/benchmarks/vibe-code
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,webdev-arena,Text Arena · Coding,coding,a93b918a1a998e2c3f9454260b7ca0ad9bdb1cb19a97cf7ec41ece2826309fff,d8ad86118f77d58db75f14ecaa73f44fd0574f82e700be2b404db6b65006366a,2b9f0cbce594c67913005d67f3a3ba04b9bf59f9c186a74a7120602e3306f4dc,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1395.6212231615589,1396,exact,canonical-source,profile_01f55deedf8de2a463a745ff,medium,medium,true,fixed_effort,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,vending-bench-2,Vending-Bench 2,multi-turn-agent,192b502afef0b6480f37700041c06c2bb2a22ecb10051d10395de1eac7c5e89f,ab877d14ba4a4b3f6a4a4fe2680927867c451e624165f39403d9b68f42004269,b2fc64da721ffa54d136f2e30205fd93a251d487411373d4ee28a27083d42d9d,release_143d5eb012c28bfa29646c4c,Vending-Bench 2 current leaderboard,one-simulated-year-mean-final-balance,mean_final_balance_usd,USD,true,1473.432,1473.43,exact,canonical-source,profile_cd54c0156f4acd0d4d181e31,reported,unknown,,unknown,,false,2026-09-02T09:26:59Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-vending-bench-2,Epoch AI mirror · Vending-Bench 2,https://epoch.ai/benchmarks/vending-bench-2
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,8f30bfa4e32dbe85773a3c2ef50565ff7d135cccb42c9cf9e314eef8ef7a21c7,c68fe87f3f9f299aee31757dff57fb40083c5e1fe694f1b8a5495597eb238754,b7aaa2498276050a4a4bfd46e4fc628034454fa6aafc3a2397d0f4e61bcf8229,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,833.9,834,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-10-31T04:32:28Z,fresh,0,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
openai/gpt-5-1,GPT 5.1,OpenAI,closed,98,weirdml,WeirdML,coding,1cb557284cca5802dafff3b02871cc031f4dd051f1946b9f6893b881bafc7e47,db4df31bc3468e838bc271f6716f5a964d2967748678a908aa80e9147814b5d2,b6a4d86486624be5cca6ad8599ec08e1124238eb955a20cfbad728c24bd3796c,release_ed6368cba447631077bb553f,WeirdML v2 current leaderboard,weirdml-v2-task-set,accuracy,percent,true,60.77,60.8,exact,canonical-source,profile_30f1bfb383c02c782db74a79,high,high,true,fixed_effort,,false,2026-09-02T09:26:51Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-weirdml,Epoch AI mirror · WeirdML,https://epoch.ai/benchmarks/weirdml
google/gemma-4-31b,Gemma 4 31B,Google,open,99,arena-text,Arena Text,preference,86aea1fca4eb86a9ce2aed76c636f4f5c58798758048dd5327362dad05c6fea3,2110659971275710a2cdc28dd4d8441b312fab8e7397c41349e1c3486d8e1824,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1452.2892381987065,1452,exact,canonical-source,profile_2a1871ad8f8b32a3d2496758,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
google/gemma-4-31b,Gemma 4 31B,Google,open,99,hle-text,Humanity's Last Exam · Text-only,expert-knowledge,b7253125f049e261eefebbac239d0cf90be3605356f662b49205a48bfa04c785,2098aed4017bebd26ae1fddc25e519b325d3a04760a38e49746c2dfb6882e0be,3b696e35852ae8aeca6f62ba64b68adde47e8a9657c04c50073322ab819e730c,release_def1fd244db3612317ca0108,HLE text-only (Artificial Analysis),text-only-2158-no-tools,accuracy,percent,true,23.632994,23.6,exact,canonical-source,profile_19bb152ad555330fda392da5,adaptive,adaptive,true,adaptive,,false,2026-09-04T00:52:24Z,fresh,58,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemma-4-31b,Gemma 4 31B,Google,open,99,simpleqa-verified,SimpleQA Verified,factuality,7fce75697ae036d6fbf22299042bf87a2e97fcd6873217c0cfc32a830b8dbf37,e613ea24e26a9c6d3025330ccd588f61dd4c343add3fa7b3563930238ce8bd71,80ee7690245a67a4e25f4db6369059e8b8b8fe3bad89432648ac923f3b47c398,release_ccacd6f679c9142b534364b4,task-v7,verified,accuracy,percent,true,10.683760683760683,10.7,other-supplemental,supplemental-source,profile_d495975af2f81492de3a7940,reported,unknown,,unknown,,false,2026-06-01T11:15:20.646874Z,fresh,152,1,A,1,A,0,A+,official-benchmark-platform,0,official-benchmark-platform,simpleqa-verified,SimpleQA Verified,https://www.kaggle.com/api/v1/benchmarks/deepmind/simpleqa-verified/leaderboard
google/gemma-4-31b,Gemma 4 31B,Google,open,99,gpqa-diamond,GPQA Diamond,science,0258091482e3712468d5a7ce2dace3dfb71f0b5e6bea5a9147faa29c532ce2d2,1437bf75f3ce05b9467adc8add10dd652b9e5109a133d9c3cf3b8c45cd316e69,077fb0f0faab782835c9f7bfe28335643bb93776086bf93c619e68f7e2781fb6,release_f83cfc2a9e19b8d654d7b58e,GPQA Diamond (Artificial Analysis),diamond-198-zero-shot-5-repeats,accuracy,percent,true,85.656566,85.7,other-supplemental,supplemental-source,profile_19bb152ad555330fda392da5,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemma-4-31b,Gemma 4 31B,Google,open,99,mmmu-pro,MMMU-Pro,multimodal,bb916308dcfd798993d968a74b0b67c88ff6bc278e787bd08868a870cd2ef68d,1add8432d15b90efce68a7907fe9707fcccd79ce811fd1615afd74cb22de86f4,8c5883ee3cfeab07c88669ed8d7bc36ab78c038c79412e35b4694c92c6776797,release_46396ea50bd4792fc96cf001,MMMU-Pro (Artificial Analysis),aa-mmmu-pro,accuracy,percent,true,73.410405,73.4,other-supplemental,supplemental-source,profile_19bb152ad555330fda392da5,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemma-4-31b,Gemma 4 31B,Google,open,99,critpt,CritPt,science,cd9d46e1d9af609be3d03ec63dc9ebad06c33b482869487068ddcef4421d863d,a098cd1640d608908838e6aafb9a1b8dff4544a87b30252b586b7323f706d024,bf184735759ce0ec8d6d8a0b14c6c8cc329f4d58b343218f782a125f7345b2a2,release_a71cc2d98bd454276f8e6ed8,CritPt (Artificial Analysis),70-challenges-official-grading,accuracy,percent,true,1.428571,1.4,exact,canonical-source,profile_19bb152ad555330fda392da5,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemma-4-31b,Gemma 4 31B,Google,open,99,swe-rebench-v2,SWE-rebench v2,coding-agent,929bd5982ea3d47d29cc9bb1c2c8461e12257f6b7edd487f5b820ed41052bd1b,7fb574026c65a7c2dd145513da22ab30597fc39d6ee1375c86ef6fc6b323e56b,716d95b37f928778ed736abc0cc74aaf85e769a20276a1ac7fbf52aecae953dd,release_1f8553cc12d8235844f1a578,rolling-v2,2026-04-15/2026-05-15,resolved_rate,percent,true,16.410256410256412,16.41,other-supplemental,supplemental-post-release-window,profile_95be02b559104aaf55501db0,reported,unknown,,unknown,,false,2026-05-15T00:00:00Z,fresh,170,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
google/gemma-4-31b,Gemma 4 31B,Google,open,99,scicode,SciCode,coding,9403844b3ec158662a799c2a9f37244dac80ba6f62dc91f904dd717c338a79cf,bf6cdf859bf19d922ae01cce0b64128bc87b92e7172c4791369597a4808a700c,bc01027d39523f00a61ff825a56baa1605c8f518dd718bba3bd0fceafe2c05b4,release_de40c66bd40a7b5483b40f9a,SciCode v1.0.1 (Artificial Analysis),scientific-research-coding-subproblems,subproblem_accuracy,percent,true,45.486111,45.5,other-supplemental,supplemental-source,profile_19bb152ad555330fda392da5,adaptive,adaptive,true,adaptive,,false,2026-09-05T17:21:27Z,fresh,56,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemma-4-31b,Gemma 4 31B,Google,open,99,webdev-arena,Text Arena · Coding,coding,c8d0c0a1f3e09c8ee633624df33b28a2692dd43e59313a46ef2338ce1eaaaec9,d83b532e70e8319cb250bc802515bfa174b674000b6b80e932bc1ec062fdf6c5,d9ab878fcc878ac812cc8fd92c382609692e2b534ea037af912dd4f876d6dcfe,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1368.6624108190852,1369,exact,canonical-source,profile_5156ce72059ef2337abe25ad,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
google/gemma-4-31b,Gemma 4 31B,Google,open,99,terminal-bench-2.1,Terminal-Bench 2.1,terminal-agent,ad885264f77518f7b1b6fada5377b9894026e0d3edd4c4a356c07896fc723ec9,dee374054f2358005bb1a6660380071aeb1ca31b80e008e1c3d8647d2a4d5094,17028b2a7321b4d1be03b6756ded6e84f7faf259b747db3071110df727184327,release_c53c830678318dc1763793c8,Terminal-Bench 2.1 (Artificial Analysis),terminus-2-89-tasks,pass_at_1,percent,true,43.445693,43.4,other-supplemental,supplemental-source,profile_19bb152ad555330fda392da5,adaptive,adaptive,true,adaptive,,false,2026-09-02T13:33:07Z,fresh,59,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemma-4-31b,Gemma 4 31B,Google,open,99,terminal-bench-4,Terminal-Bench 4.0,terminal-agent,889f19bc21527a943215164c273f14c14794dd4586a78ca8b7a1a695f11511fe,bbac00547575b721314e0a95821c96ed5a47accef34c83953bb2214443c14e1c,7ba5119d29b93344b7a8f6da0817d11aa98b8ccebf254c12ef9825cf15af5b3a,release_b7d9ab744ffac0cb856e114d,Terminal-Bench 4.0 (Artificial Analysis),mini-swe-agent-66-tasks-500-steps,pass_at_1,percent,true,0.0,0.0,other-supplemental,supplemental-source,profile_19bb152ad555330fda392da5,adaptive,adaptive,true,adaptive,,false,2026-10-21T23:21:17Z,fresh,10,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemma-4-31b,Gemma 4 31B,Google,open,99,tau3-banking-aa,τ³-Banking · Artificial Analysis harness,multi-turn-agent,1fa3c0cfe05a8a2838fd4440b289329bf51d685a1b4d725608f6eab5f3a53035,3f555090c12b141c2ffd74e65e706f005b466baebc654c71548860d9505c8cc9,726cb0966b46b174339f6cf4f1aaef8a6160b2984ccd900ccdd9c3f3b084c43e,release_786d72c2e6dda7b72acc4040,τ³-Banking (Artificial Analysis),banking-97-tasks-gpt-5-4-mini-simulator,pass_at_1,percent,true,14.845361,14.8,exact,canonical-source,profile_19bb152ad555330fda392da5,adaptive,adaptive,true,adaptive,,false,2026-09-04T09:03:13Z,fresh,57,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/leaderboards/models
google/gemma-4-31b,Gemma 4 31B,Google,open,99,gdpval-aa-v2.1,GDPval-AA v2.1,professional-work,3429e7b1eff53c331a19ad5f402b4fa366b258bccf30428011ca229659c6722d,d572c9507b03fcef28c7e5c5ec5edb4cc8d17eebae1b7cee6254e6e09255af3d,938eed1e68edafc7a732bfad1af0285e449fdba9d99fa3fe2e2d11aa7e40f617,release_2b38a48aaf3cb40aca8959b1,"GDPval-AA v2.1 (Crowd-BT Elo, DeepSeek V4.1 Flash (max) at 1600)",220-tasks-elo-v2-1,elo,points,true,626.77,627,exact,canonical-source,profile_19bb152ad555330fda392da5,adaptive,adaptive,true,adaptive,,false,2026-10-29T17:25:13Z,fresh,2,1,A,1,A,1,A,independent-evaluation,6,independent-evaluation,artificial-analysis,Artificial Analysis · Intelligence Index evaluations,https://artificialanalysis.ai/evaluations/gdpval-aa
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,epoch-eci,Epoch Capabilities Index,capability-index,70c9d431223f6ba1702906365ae3071bc66860adde5acba73a57ff9d287675ff,00da902e41f51f32bc647b2af03fb15004f00ef5cc388f7a21b8e86d8d82c016,14226ff6d4f36f6d9b6090b420212bce2e20f6b6260c7e70d44f9d3063be385c,release_139e6a0a9905605b7ed14674,Epoch ECI snapshot 2026-11-01,official-composite-fit,eci,index_points,true,143.93,143.93,exact,canonical-source,profile_887d48f324f77d968eaecc1d,reported,unknown,,unknown,,false,2026-11-01,fresh,0,1,A,1,A,0,A+,official-benchmark-platform,5,official-benchmark-platform,epoch-eci,Epoch Capabilities Index,https://epoch.ai/eci
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,arena-text,Arena Text,preference,72e0c0ccb95d4237c31a4d257ee3ef5ce25a35a65c5537bc3b92afff28d5a607,4a00f5535e91d873e919708539862c3f160986d6e3e874673ccbe8ebf2980286,ffbf0922679e4aa0b4bec1c541f932b1c512061f56ecc5ebc569b49807caa020,release_482c53e63fe96f7e715bb0f5,2026-10-28,overall-style-control,arena_score,points,true,1448.0134195918026,1448,exact,canonical-source,profile_aa1f16a5a40da24207af00f4,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-text,Arena Text,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/text_style_control/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,gpqa-diamond,GPQA Diamond,science,e31e32663a242cac00eb00ed785e1f65436621148e12c2b868b95eb0deaff73c,8cb8f3536ddc1bf09d87d525f1b7d1ed198749f77148cb5314cd3f9effd5d20a,d0e5c8fe1e46628e66d220e813b80a61c67420010f87ce00aaad4d7aac448d42,release_aa2f2a6181ec4d1bc71a32f8,GPQA Diamond v1 · Vals dual-prompt,diamond-198-zero-and-five-shot-cot,accuracy,percent,true,76.263,76.26,exact,canonical-source,profile_99dd806d3a14424dd2a38e68,adaptive,adaptive,true,adaptive,,false,2026-08-30T23:35:12Z,fresh,62,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-gpqa-diamond,Vals AI GPQA Diamond,https://www.vals.ai/benchmarks/gpqa
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,mmlu-pro,MMLU-Pro,knowledge,f2512b4b94e058422aadc8361a1f21e7e1d622042a10aa99e845774cca603bd9,68199229cb5c90078b16ff6989933af6b824f5e92dfbb0693d41b36deadbbc2c,65862a584a4422b9d9ef6c99a9d4fef3b04b48a3e7c375f14406331208167a33,release_187c8201a49a6385bb527d7d,Vals MMLU-Pro v1,vals-mmlu-pro-overall,accuracy,percent,true,87.924,87.9,exact,canonical-source,profile_99dd806d3a14424dd2a38e68,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:37Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmlu-pro,Vals AI · MMLU-Pro,https://www.vals.ai/benchmarks/mmlu_pro
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,frontiermath,FrontierMath T1–3,mathematics,73d80f61332f67f221690497baf00e9a255c88726901d12384669bc86a50e3d9,5c06bd1da8dc65637d368f43918015a17243ca5236184fa131a6aede0d3faf81,5b3536a86ff25f5a7597b213ff2cdf948aea0b0996d8daccf3472da9e0577a56,release_fd01c0ab187a88391c43ea9f,2.0.0,private-285-of-295-problem-tiers-1-3-set,mean_score,percent,true,12.631578947368421,12.6,exact,canonical-source,profile_01129548b25adbb9776ec287,thinking-32k,unknown,,token_budget,32000,false,2026-06-11T21:15:07.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath,Epoch AI · FrontierMath Tiers 1–3 v2,https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,frontiermath-tier-4,FrontierMath T4,mathematics,5ea1f355eaa80b08ae1ae1426cdf49266c9d9b65cac51bca91a5fbeeac260b71,34c9f7d5feface035ef6a54552b5116d139f70d7e68d53c809242bad9f61f4e6,ae135bcb39223edd49e0623f9c0ce7bc8c33cf3ee2ca71e01b0b431feced8c16,release_4d025b5eeb2f6b23fd7b19cb,2.0.0,private-41-of-43-problem-tier-4-set,mean_score,percent,true,2.4390243902439024,2.4,exact,canonical-source,profile_539047413da7106e5111873d,thinking-32k,unknown,,token_budget,32000,false,2026-06-11T21:15:07.000Z,fresh,142,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-frontiermath-tier-4,Epoch AI · FrontierMath Tier 4 v2,https://epoch.ai/benchmarks/frontiermath-tier-4-v2
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,mmmu-pro,MMMU-Pro,multimodal,eedaddb7adc349bd6dd5afa8ee94e4f4fe0d33318e5521ba5d773bf5920c32aa,e936072964a06c00397553515f6b1935d52f185754768fae3b1318e63f83ce8a,1dc13b2caaf2b852bef31bf6ea73b3bab0bdc1b0693a254502b57a105da24c42,release_c4828ed5eb385de228d39263,Vals MMMU-Pro v1,vals-mmmu-pro-overall,accuracy,percent,true,77.514,77.5,exact,canonical-source,profile_99dd806d3a14424dd2a38e68,adaptive,adaptive,true,adaptive,,false,2026-10-08T22:23:22Z,fresh,23,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-mmmu-pro,Vals AI · MMMU-Pro,https://www.vals.ai/benchmarks/mmmu
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,simplebench,SimpleBench,reasoning,13f05786427f998de535c10418ee499b5b1db5fae5ef8d36177cb832f6e1df96,6b2c522046cad1c232c0bebc6e1cce4f8d18e4538238d206f994be990979c780,56504c70f5d45eef26f9b98e3fb6f9f02f5f3a2076da3de5c1556a7ca296352b,release_af5e4653e36272fb4c37bce2,SimpleBench current leaderboard,official-set-avg-at-5,accuracy_avg_at_5,percent,true,60.0,60.0,exact,canonical-source,profile_dfc3eed7d38c3a0d0a68268a,reported,unknown,,unknown,,false,2026-09-02T09:26:45Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-simplebench,Epoch AI mirror · SimpleBench,https://epoch.ai/benchmarks/simplebench
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,swe-bench-verified,SWE-bench Verified,coding-agent,eb9348120b6b5399bc7f2916104f0a8ece71cf01f039a6619b6c6f8b94a90c3d,391888eea54099d97dece71c99690601bb5a1cda21cb09c7bbb6c0a3a29cd99b,a2b04faf2230205e13b4c801da0beea2e8239f0cb8ddc3954f0007d9771dd68b,release_ab91f0254cf52d5b135599b2,Epoch independent runs,verified-484,resolved_rate,percent,true,73.34710743801654,73.3,other-supplemental,supplemental-source,profile_832ee21d16c8e7eaa53c7f7c,reported,unknown,,unknown,,false,2026-02-11T21:33:53.502Z,stale,262,1,A,1,A,1,A,independent-evaluation,1,independent-evaluation,epoch-swe-bench-verified,Epoch AI SWE-bench Verified,https://epoch.ai/data/benchmark_data.zip
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,swe-rebench-v2,SWE-rebench v2,coding-agent,c59ed59dc77980cb9cab597acd02da0e58240ba1e5d0a74df6415e4b36bc1a66,0bcd0ff83cf8dbf795666b3b01cd87a0be2bc0433dd24521f2ecc20138ad87fb,122b54f28b6294368d79a290bafcab91ec40542b45f7262adfc41a652c854b80,release_e8fc75e5c8482265dbca3d4c,rolling-v2,2025-09-01/2025-12-01,resolved_rate,percent,true,43.46193271829281,43.46,other-supplemental,supplemental-prior-window,profile_4efca96ee4b3e901a2b47082,reported,unknown,,unknown,,false,2025-12-01T00:00:00Z,stale,335,0,A+,0,A+,0,A+,official-leaderboard,2,official-leaderboard,swe-rebench,SWE-rebench,https://swe-rebench.com/
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,livecodebench,LiveCodeBench,coding,3c694f1bee4db4a8476981048f361cfdc1ca70954a659f4a213034e5ebfd5f90,b98cb153d75b2486bff8be384096e06862f3815a518b77a122f8c5d27e9fbd20,a93d4c3dde012a7d2fb80c568e83c4c7745915c2757016dd2d0fb1a39ff361da,release_a2492ed53dd6d8c4d33d5ae5,Vals LiveCodeBench v1,vals-livecodebench-overall,pass_at_1,percent,true,66.456,66.5,exact,canonical-source,profile_99dd806d3a14424dd2a38e68,adaptive,adaptive,true,adaptive,,false,2026-09-02T09:40:32Z,fresh,59,2,B,2,B,1,A,independent-evaluation,6,independent-evaluation,vals-livecodebench,Vals AI · LiveCodeBench,https://www.vals.ai/benchmarks/lcb
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,webdev-arena,Text Arena · Coding,coding,7b194f6079c9005fc070d934cda6bbfbdccd889862cea27a67302d7356339204,c40e3ea20f0d9467d1d15468eef16aa07ce4d64a6917244d84716fbac0caef40,d818bf781585955b264603ddf7816ce1a68a99d200a5e6e6c703d1ecdd0823be,release_254ac63a5db19f0802d10f9f,2026-10-28,overall-code-arena,arena_score,points,true,1390.1558311618264,1390,exact,canonical-source,profile_0447edc48ed9dd787c5b6bc9,reported,unknown,,unknown,,false,2026-10-28,fresh,4,1,A,1,A,0,A+,official-leaderboard,4,official-leaderboard,arena-webdev,LMArena WebDev Arena,https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/323062f70843f40e3f739865c3fe9d806126e412/webdev/latest-00000-of-00001.parquet?download=true
anthropic/claude-opus-4-1,Claude Opus 4.1,Anthropic,closed,100,gdpval,GDPval,professional-work,d1fe3e2c4bf2762a2383f9414bf420890571b0c430274fee462ef793796c5ca6,955814e05796e63726cdea69007c67fd84c8ecabc11973ba8519e95ddf3ff289,2b9517ae4828c51845ede7bca6aadbc8eb35e299be4327218a6131dee37ed9fb,release_730527fa346470703bd19e21,GDPval official leaderboard,expert-comparison-win-rate,win_rate,percent,true,43.6,43.6,exact,canonical-source,profile_73ca51c1241eb639e187e8eb,reported,unknown,,unknown,,false,2026-09-02T09:27:05Z,fresh,59,2,B,2,B,3,B,official-secondary-verification,6,official-secondary-verification,epoch-gdpval,Epoch AI mirror · GDPval,https://epoch.ai/benchmarks/gdpval