Buckets:
celinah/hf-cli-agent-benchmark / T07-find-coding-models /claude-code-sonnet /cli-skill /cli_skill-r0-3faf62.jsonl
| {"parentUuid": null, "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": "Find up to 8 trending text-generation models for coding, released in 2025-2026, with fewer than 70B parameters and suitable for local deployment. Exclude the mainstream Llama, Mistral and Qwen base series. Return the model IDs with their parameter counts."}, "uuid": "bd8b72cc-0914-5f4b-a3b2-2a8c80bffd64", "timestamp": "2026-06-01T12:00:00.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "bd8b72cc-0914-5f4b-a3b2-2a8c80bffd64", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "thinking", "thinking": "The user wants to find trending text-generation models for coding, released in 2025-2026, with fewer than 70B parameters, suitable for local deployment, excluding mainstream Llama, Mistral, and Qwen b", "signature": ""}]}, "uuid": "d2975799-b99f-57a1-86b6-215b93d6b061", "timestamp": "2026-06-01T12:00:01.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "d2975799-b99f-57a1-86b6-215b93d6b061", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "text", "text": "I'll search for trending text-generation models focused on coding using the HF CLI."}]}, "uuid": "727c2f8a-761a-5442-b72f-60f54a4cbbf4", "timestamp": "2026-06-01T12:00:02.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "727c2f8a-761a-5442-b72f-60f54a4cbbf4", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_c88cb40be01d5d0bbcc5d7", "name": "Bash", "input": {"raw": "{'command': 'hf models list --filter \"text-generation\" --sort trending_score --limit 50 --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\ndata = json.load(sys.stdin)\\nresults = []\\nfor m in data:\\n mid = m.get(\\'id\\',\\'\\')\\n # exclude mainstream base series\\n low = mid.lower()\\n if any(x in low for x in [\\'llama\\', \\'mistral\\', \\'qwen\\']):\\n continue\\n # only 2025-2026\\n created = m.get(\\'createdAt\\',\\'\\') or m.get(\\'created_at\\',\\'\\') or \\'\\'\\n last_mod = m.get(\\'lastModified\\',\\'\\') or m.get(\\'last_modified\\',\\'\\') or \\'\\'\\n date_str = created or last"}}]}, "uuid": "2819b26a-c5dc-535a-9825-0534928520a3", "timestamp": "2026-06-01T12:00:03.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "2819b26a-c5dc-535a-9825-0534928520a3", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_c88cb40be01d5d0bbcc5d7", "content": "[\n {\n \"id\": \"LiquidAI/LFM2.5-8B-A1B\",\n \"created_at\": \"2026-05-28T09:43:54+00:00\",\n \"downloads\": 60171,\n \"library_name\": \"transformers\",\n \"likes\": 457,\n \"pipeline_tag\": \"text-generation\",\n \"private\": false,\n \"tags\": [\n \"transformers\",\n \"safetensors\",\n \"lfm2_moe\",\n \"text-generation\",\n \"liquid\",\n \"lfm2.5\",\n \"edge\",\n \"conversational\",\n \"en\",\n \"ar\",\n \"zh\",\n \"fr\",\n \"de\",\n \"ja\",\n \"ko\",\n \"es\",\n \"pt\",\n \"arxiv:2511.23404\",\n \"base_model:LiquidAI/LFM2.5-8B-A1B-Base\",\n \"base_model:fi"}]}, "uuid": "fd4b1b57-a9f7-5f84-841d-fbff9edbf986", "timestamp": "2026-06-01T12:00:04.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "fd4b1b57-a9f7-5f84-841d-fbff9edbf986", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_141965607abd5d00892660", "name": "Bash", "input": {"raw": "{'command': 'hf models list --filter \"text-generation,code\" --sort trending_score --limit 50 --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\ndata = json.load(sys.stdin)\\nresults = []\\nfor m in data:\\n mid = m.get(\\'id\\',\\'\\')\\n low = mid.lower()\\n if any(x in low for x in [\\'llama\\', \\'mistral\\', \\'qwen\\']):\\n continue\\n created = m.get(\\'createdAt\\',\\'\\') or m.get(\\'created_at\\',\\'\\') or \\'\\'\\n last_mod = m.get(\\'lastModified\\',\\'\\') or m.get(\\'last_modified\\',\\'\\') or \\'\\'\\n date_str = created or last_mod\\n if not (date_str.startswith(\\'2025\\') or date"}}]}, "uuid": "e581272d-565f-582e-8743-86dd04973e27", "timestamp": "2026-06-01T12:00:05.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "e581272d-565f-582e-8743-86dd04973e27", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_141965607abd5d00892660", "content": "[\n {\n \"id\": \"XiaomiMiMo/MiMo-V2.5-Pro\",\n \"trending\": 20,\n \"created\": \"2026-04-27\",\n \"likes\": 579\n },\n {\n \"id\": \"jedisct1/MiMo-V2.5-coder-Q2\",\n \"trending\": 17,\n \"created\": \"2026-05-24\",\n \"likes\": 71\n },\n {\n \"id\": \"RJ000/Mellum2-12B-A2.5B-Thinking-GGUF\",\n \"trending\": 6,\n \"created\": \"2026-06-02\",\n \"likes\": 6\n },\n {\n \"id\": \"EssentialAI/rnj-1.5-instruct\",\n \"trending\": 5,\n \"created\": \"2026-04-22\",\n \"likes\": 12\n },\n {\n \"id\": \"hiiamboris/Darwin-28B-Coder-Q4_K_M-GGUF\",\n \"trending\": 5,\n \"created\": \"2026-05-21\",\n \"likes\": 5\n },\n {\n "}]}, "uuid": "f56b1273-aeba-5788-8aaa-f41f129e498b", "timestamp": "2026-06-01T12:00:06.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "f56b1273-aeba-5788-8aaa-f41f129e498b", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_90e64ca031035397bf289d", "name": "Bash", "input": {"raw": "{'command': '# Search specifically for coding models with various tags\\nhf models list --search \"coder\" --filter \"text-generation\" --sort trending_score --limit 50 --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\ndata = json.load(sys.stdin)\\nresults = []\\nfor m in data:\\n mid = m.get(\\'id\\',\\'\\')\\n low = mid.lower()\\n if any(x in low for x in [\\'llama\\', \\'mistral\\', \\'qwen\\']):\\n continue\\n created = m.get(\\'createdAt\\',\\'\\') or m.get(\\'created_at\\',\\'\\') or \\'\\'\\n date_str = created\\n if not (date_str.startswith(\\'2025\\') or date_str.startswith(\\'2026\\')):"}}]}, "uuid": "20ef3505-7d35-56b6-b884-433aacc145d0", "timestamp": "2026-06-01T12:00:07.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "20ef3505-7d35-56b6-b884-433aacc145d0", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_90e64ca031035397bf289d", "content": "[\n {\n \"id\": \"jedisct1/MiMo-V2.5-coder-Q2\",\n \"trending\": 17,\n \"created\": \"2026-05-24\",\n \"likes\": 71\n },\n {\n \"id\": \"Jackrong/Qwopus3.5-4B-Coder\",\n \"trending\": 10,\n \"created\": \"2026-05-26\",\n \"likes\": 10\n },\n {\n \"id\": \"hiiamboris/Darwin-28B-Coder-Q4_K_M-GGUF\",\n \"trending\": 5,\n \"created\": \"2026-05-21\",\n \"likes\": 5\n },\n {\n \"id\": \"k0valik/Darwin-28B-Coder-IQ4_XS12.9GiB-GGUF\",\n \"trending\": 5,\n \"created\": \"2026-05-21\",\n \"likes\": 5\n },\n {\n \"id\": \"jedisct1/MiMo-V2.5-coder-Q2-v2\",\n \"trending\": 5,\n \"created\": \"2026-05-25\",\n \"likes\": 7\n }"}]}, "uuid": "33c44d53-7ce0-5368-94a2-f144215c0fa2", "timestamp": "2026-06-01T12:00:08.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "33c44d53-7ce0-5368-94a2-f144215c0fa2", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_5bcdf61f4d2c585894748f", "name": "Bash", "input": {"raw": "{'command': '# Search for well-known coding models by name\\nfor model in \"deepseek-ai/DeepSeek-Coder-V2\" \"deepseek-ai/DeepSeek-R2\" \"XiaomiMiMo/MiMo-V2.5-Pro\" \"JetBrains/Mellum2-12B-A2.5B-Thinking\" \"microsoft/Phi-4-mini-instruct\" \"microsoft/Phi-4-reasoning\" \"JetBrains/Mellum-4b-base\" \"google/gemma-3\" \"codefuse-ai\" \"TeamACE/ToolACE\" \"ByteDance-Seed/Seed-Coder-8B-Base\"; do\\n echo \"--- $model ---\"\\ndone\\n\\n# Let\\'s get models from key coding model families\\nhf models list --author \"deepseek-ai\" --filter \"text-generation\" --sort trending_score --limit 20 --format json 2>/dev/null | python3 -c \"\\ni"}}]}, "uuid": "94d52275-0691-5de9-8adc-d01e3a681d02", "timestamp": "2026-06-01T12:00:09.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "94d52275-0691-5de9-8adc-d01e3a681d02", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_5bcdf61f4d2c585894748f", "content": "--- deepseek-ai/DeepSeek-Coder-V2 ---\n--- deepseek-ai/DeepSeek-R2 ---\n--- XiaomiMiMo/MiMo-V2.5-Pro ---\n--- JetBrains/Mellum2-12B-A2.5B-Thinking ---\n--- microsoft/Phi-4-mini-instruct ---\n--- microsoft/Phi-4-reasoning ---\n--- JetBrains/Mellum-4b-base ---\n--- google/gemma-3 ---\n--- codefuse-ai ---\n--- TeamACE/ToolACE ---\n--- ByteDance-Seed/Seed-Coder-8B-Base ---\ndeepseek-ai/DeepSeek-V4-Pro | 2026-04-22 | trending=186 | likes=4588\ndeepseek-ai/DeepSeek-R1 | 2025-01-20 | trending=15 | likes=13364\ndeepseek-ai/DeepSeek-R1-Distill-Qwen-7B | 2025-01-20 | trending=4 | likes=841\ndeepseek-ai/DeepSeek-V3-03"}]}, "uuid": "fe865e2e-c9d7-5185-9ec1-07d07304a268", "timestamp": "2026-06-01T12:00:10.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "fe865e2e-c9d7-5185-9ec1-07d07304a268", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_505e666ea1f758ab9f825a", "name": "Bash", "input": {"raw": "{'command': '# Get model info with parameter counts for top candidates\\n# Let\\'s look at key coding models\\nfor model_id in \"deepseek-ai/DeepSeek-R1-Distill-Qwen-7B\" \"XiaomiMiMo/MiMo-V2.5-Pro\" \"JetBrains/Mellum2-12B-A2.5B-Thinking\" \"microsoft/Phi-4-mini-instruct\" \"JetBrains/Mellum-4b-base\" \"ByteDance-Seed/Seed-Coder-8B-Base\" \"Tesslate/OmniCoder-9B\" \"openbmb/MiniCPM5-1B\"; do\\n echo \"=== $model_id ===\"\\n hf models info \"$model_id\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\nd = json.load(sys.stdin)\\nprint(\\' created:\\', d.get(\\'created_at\\',\\'\\')[:10])\\nprint(\\' trending:\\', "}}]}, "uuid": "8e68f579-f931-5f70-8346-e97777a76568", "timestamp": "2026-06-01T12:00:11.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "8e68f579-f931-5f70-8346-e97777a76568", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_505e666ea1f758ab9f825a", "content": "=== deepseek-ai/DeepSeek-R1-Distill-Qwen-7B ===\n created: 2025-01-20\n trending: None\n likes: 841\n parameters: {'parameters': {'BF16': 7615616512}, 'total': 7615616512}\n num_params: None\n=== XiaomiMiMo/MiMo-V2.5-Pro ===\n created: 2026-04-27\n trending: None\n likes: 579\n parameters: {'parameters': {'F32': 224695296, 'BF16': 9450733440, 'F8_E4M3': 1013569290240}, 'total': 1023244718976}\n num_params: None\n=== JetBrains/Mellum2-12B-A2.5B-Thinking ===\n created: 2026-05-26\n trending: None\n likes: 160\n parameters: {'parameters': {'BF16': 12149923072}, 'total': 12149923072}\n num_params: N"}]}, "uuid": "9bc5e03b-43dd-5cd2-b211-6664a40c51b4", "timestamp": "2026-06-01T12:00:12.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "9bc5e03b-43dd-5cd2-b211-6664a40c51b4", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_5ff117d3f47453739156d6", "name": "Bash", "input": {"raw": "{'command': '# MiMo-V2.5-Pro has 1T+ params - too large. Let\\'s check DeepSeek-R1 distill models, \\n# Phi-4 reasoning, and look for more targeted coding models\\n# Also check specific coding-focused model families\\n\\n# Search for more coding-specific trending models\\nhf models list --search \"coding\" --filter \"text-generation\" --sort trending_score --limit 30 --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\ndata = json.load(sys.stdin)\\nfor m in data:\\n mid = m.get(\\'id\\',\\'\\')\\n low = mid.lower()\\n if any(x in low for x in [\\'llama\\', \\'mistral\\', \\'qwen\\']):\\n continu"}}]}, "uuid": "edbabfce-9b81-59b3-819c-e3e08494f7dc", "timestamp": "2026-06-01T12:00:13.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "edbabfce-9b81-59b3-819c-e3e08494f7dc", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_5ff117d3f47453739156d6", "content": "---"}]}, "uuid": "1ab7aff0-71b4-5fc6-afdb-f31cda66f303", "timestamp": "2026-06-01T12:00:14.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "1ab7aff0-71b4-5fc6-afdb-f31cda66f303", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_f4fde2b35812511ea9e3f1", "name": "Bash", "input": {"raw": "{'command': '# Let\\'s look at trending models more broadly and check for coding ones\\nhf models list --filter \"text-generation\" --sort trending_score --limit 100 --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\ndata = json.load(sys.stdin)\\nresults = []\\nfor m in data:\\n mid = m.get(\\'id\\',\\'\\')\\n low = mid.lower()\\n # exclude mainstream base series (but allow derived/distilled)\\n if any(x in low for x in [\\'llama\\', \\'mistral\\']):\\n continue\\n # Allow qwen-based coding fine-tunes but exclude qwen base\\n created = m.get(\\'created_at\\',\\'\\')[:10]\\n if not ("}}]}, "uuid": "ea06c29b-23bf-5b86-969a-3ba2f2e69e02", "timestamp": "2026-06-01T12:00:15.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "ea06c29b-23bf-5b86-969a-3ba2f2e69e02", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_f4fde2b35812511ea9e3f1", "content": "LiquidAI/LFM2.5-8B-A1B | 2026-05-28 | ?B | trending= 446 | code=False\nopenbmb/MiniCPM5-1B | 2026-05-21 | ?B | trending= 348 | code=True\nstepfun-ai/Step-3.7-Flash | 2026-05-23 | ?B | trending= 218 | code=True\ndeepseek-ai/DeepSeek-V4-Pro | 2026-04-22 | ?B | trending= 186 | code=False\nLiquidAI/LFM2.5-8B-A1B-GGUF | 2026-05-24 | ?B | trending= 161 | code=False\nJetBrains/Mellum2-12B-A2.5B-Thin"}]}, "uuid": "1054a554-184b-53bc-8ea1-743802d0530e", "timestamp": "2026-06-01T12:00:16.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "1054a554-184b-53bc-8ea1-743802d0530e", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_881296fa524c5980ad6218", "name": "Bash", "input": {"raw": "{'command': '# Now get detailed info on top candidates for coding + parameter count\\n# Focus on known coding models: DeepSeek-Coder variants, Phi-4, JetBrains Mellum, \\n# Seed-Coder, Darwin, MiMo (smaller)\\n\\nfor model_id in \"deepseek-ai/DeepSeek-R1-Distill-Qwen-7B\" \"microsoft/Phi-4-mini-instruct\" \"microsoft/Phi-4-reasoning\" \"JetBrains/Mellum2-12B-A2.5B-Thinking\" \"JetBrains/Mellum-4b-base\" \"ByteDance-Seed/Seed-Coder-8B-Base\" \"XiaomiMiMo/MiMo-7B-RL\" \"google/gemma-3-27b-it\" \"deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct\"; do\\n result=$(hf models info \"$model_id\" --format json 2>/dev/null)\\n if ["}}]}, "uuid": "4d446ecf-356b-5878-8e17-d316bfa66bf3", "timestamp": "2026-06-01T12:00:17.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "4d446ecf-356b-5878-8e17-d316bfa66bf3", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_881296fa524c5980ad6218", "content": "Exit code 1\ndeepseek-ai/DeepSeek-R1-Distill-Qwen-7B | 2025-01-20 | 7.6B | likes=841 | code=False\nmicrosoft/Phi-4-mini-instruct | 2025-02-19 | 3.8B | likes=750 | code=True\nmicrosoft/Phi-4-reasoning | 2025-04-09 | 14.7B | likes=225 | code=True\nJetBrains/Mellum-4b-base | 2025-04-28 | 4.0B | likes=446 | code=True\nByteDance-Seed/Seed-Coder-8B-Base | 2025-04-27 | 8.3B | likes=73 | code=True"}]}, "uuid": "c2bf536b-8dcf-5f71-bb9a-2bbc73fc1379", "timestamp": "2026-06-01T12:00:18.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "c2bf536b-8dcf-5f71-bb9a-2bbc73fc1379", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_aee92e4b49f95fa5953721", "name": "Bash", "input": {"raw": "{'command': '# Get info for more models\\nfor model_id in \"JetBrains/Mellum2-12B-A2.5B-Thinking\" \"XiaomiMiMo/MiMo-7B-RL\" \"deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct\" \"google/gemma-3-27b-it\" \"microsoft/Phi-4-reasoning-plus\" \"Tesslate/OmniCoder-9B\" \"openbmb/MiniCPM-o-3B\" \"FINAL-Bench/Darwin-60B-DUO\"; do\\n result=$(hf models info \"$model_id\" --format json 2>/dev/null)\\n if [ -n \"$result\" ]; then\\n echo \"$result\" | python3 -c \"\\nimport json, sys\\nd = json.load(sys.stdin)\\nmid = d.get(\\'id\\',\\'\\')\\ncreated = d.get(\\'created_at\\',\\'\\')[:10]\\nparams = d.get(\\'safetensors\\', {}) or {}\\ntotal = pa"}}]}, "uuid": "f683176d-23f9-5337-865e-e30f5fcb3cd8", "timestamp": "2026-06-01T12:00:19.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "f683176d-23f9-5337-865e-e30f5fcb3cd8", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_aee92e4b49f95fa5953721", "content": "microsoft/Phi-4-reasoning-plus | 2025-04-17 | 14.7B | likes=342 | code=True | trending=None\nFINAL-Bench/Darwin-60B-DUO | 2026-05-27 | 0.0B | likes=31 | code=False | trending=None"}]}, "uuid": "cd06b5b9-6636-5635-ab35-cf7a69d15d93", "timestamp": "2026-06-01T12:00:20.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "cd06b5b9-6636-5635-ab35-cf7a69d15d93", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_1064467f719c5aedaab62d", "name": "Bash", "input": {"raw": "{'command': 'hf models info \"JetBrains/Mellum2-12B-A2.5B-Thinking\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\nd = json.load(sys.stdin)\\nmid = d.get(\\'id\\',\\'\\')\\ncreated = d.get(\\'created_at\\',\\'\\')[:10]\\nparams = d.get(\\'safetensors\\', {}) or {}\\ntotal = params.get(\\'total\\', 0)\\nprint(f\\\\\"{mid} | {created} | {total/1e9:.1f}B | likes={d.get(\\'likes\\')}\\\\\")\\nprint(\\'tags:\\', d.get(\\'tags\\', [])[:10])\\n\" 2>/dev/null\\n\\nhf models info \"deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\nd = json.load(sys.stdin)\\nmid = d.get(\\'"}}]}, "uuid": "d50681b6-4bbc-5dda-b75e-21922a2b5680", "timestamp": "2026-06-01T12:00:21.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "d50681b6-4bbc-5dda-b75e-21922a2b5680", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_1064467f719c5aedaab62d", "content": "JetBrains/Mellum2-12B-A2.5B-Thinking | 2026-05-26 | 12.1B | likes=160\ntags: ['transformers', 'safetensors', 'mellum', 'text-generation', 'conversational', 'en', 'arxiv:2605.31268', 'license:apache-2.0', 'model-index', 'eval-results']\ndeepseek-ai/DeepSeek-Coder-V2-Lite-Instruct | 2024-06-14 | 15.7B | likes=603"}]}, "uuid": "6264c657-1794-5c7e-b673-d899d2063139", "timestamp": "2026-06-01T12:00:22.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "6264c657-1794-5c7e-b673-d899d2063139", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_e5ed61c11f5e539c866e5d", "name": "Bash", "input": {"raw": "{'command': '# Search for more coding-specific models from 2025-2026\\nhf models list --search \"code\" --filter \"text-generation\" --sort trending_score --limit 50 --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\ndata = json.load(sys.stdin)\\nfor m in data:\\n mid = m.get(\\'id\\',\\'\\')\\n low = mid.lower()\\n if any(x in low for x in [\\'llama\\', \\'mistral\\']):\\n continue\\n created = m.get(\\'created_at\\',\\'\\')[:10]\\n if not (created >= \\'2025-01-01\\'):\\n continue\\n params = m.get(\\'safetensors\\', {}) or {}\\n total = params.get(\\'total\\', 0)\\n if total > "}}]}, "uuid": "1cf1c796-9350-5add-a978-844be122f423", "timestamp": "2026-06-01T12:00:23.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "1cf1c796-9350-5add-a978-844be122f423", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_e5ed61c11f5e539c866e5d", "content": "Qwen/Qwen3-Coder-Next | 2026-01-30 | 0.0B | trending=21 | likes=1415\nunsloth/Qwen3-Coder-Next-GGUF | 2026-02-03 | 0.0B | trending=18 | likes=672\njedisct1/MiMo-V2.5-coder-Q2 | 2026-05-24 | 0.0B | trending=17 | likes=71\nQwen/Qwen3-Coder-30B-A3B-Instruct | 2025-07-31 | 0.0B | trending=14 | likes=1091\nunsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF | 2025-07-31 | 0.0B | trending=13 | likes=694\nJackrong/Qwopus3.5-4B-Coder | 2"}]}, "uuid": "9ef92512-d2e6-5b6e-aa05-cd3593fe0c52", "timestamp": "2026-06-01T12:00:24.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "9ef92512-d2e6-5b6e-aa05-cd3593fe0c52", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_8e953c7b427d5a46b8e3f1", "name": "Bash", "input": {"raw": "{'command': '# The Qwen series says to exclude \"base series\" - but Qwen3-Coder is a code-specific fine-tune\\n# Let me check if Qwen3-Coder is excluded. The instruction says \"Qwen base series\" \\n# so Qwen3-Coder (a coding fine-tune) might be okay. But let me look for non-Qwen alternatives.\\n\\n# Let me look at more models: Darwin-28B-Coder, MiMo-7B, Phi-4, DeepSeek-R1 distills, \\n# StarCoder2, Gemma-3, Codestral-22B\\n\\nfor model_id in \"FINAL-Bench/Darwin-28B-Coder\" \"XiaomiMiMo/MiMo-7B-RL\" \"google/gemma-3-9b-it\" \"google/gemma-3-27b-it\" \"mistralai/Codestral-22B-v0.1\" \"bigcode/starcoder2-15b\" \"deep"}}]}, "uuid": "39c1dcb0-a603-573b-b717-d7d50616b150", "timestamp": "2026-06-01T12:00:25.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "39c1dcb0-a603-573b-b717-d7d50616b150", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_8e953c7b427d5a46b8e3f1", "content": "bigcode/starcoder2-15b | 2024-02-20 | 16.0B | likes=670 | trending=None\ndeepseek-ai/DeepSeek-R1-Distill-Llama-8B | 2025-01-20 | 8.0B | likes=862 | trending=None"}]}, "uuid": "de2bce7d-4fbc-5ae7-9632-391121d36bc6", "timestamp": "2026-06-01T12:00:26.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "de2bce7d-4fbc-5ae7-9632-391121d36bc6", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_aad1fcae4c18599c999674", "name": "Bash", "input": {"raw": "{'command': '# Let me look at what\\'s actually trending and coding focused in 2025-2026\\n# Check Darwin-28B-Coder specifically\\nhf models list --search \"Darwin\" --filter \"text-generation\" --sort trending_score --limit 10 --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\ndata = json.load(sys.stdin)\\nfor m in data:\\n mid = m.get(\\'id\\',\\'\\')\\n created = m.get(\\'created_at\\',\\'\\')[:10]\\n if created >= \\'2025-01-01\\':\\n print(f\\\\\"{mid} | {created} | trending={m.get(\\'trending_score\\')} | likes={m.get(\\'likes\\')}\\\\\")\\n\" 2>/dev/null\\n\\necho \"---\"\\n\\n# Check OpenCoder, Grani"}}]}, "uuid": "ba1c98c6-1bb0-5ee7-9aaa-9e9d8b8b9880", "timestamp": "2026-06-01T12:00:27.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "ba1c98c6-1bb0-5ee7-9aaa-9e9d8b8b9880", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_aad1fcae4c18599c999674", "content": "Exit code 1\nFINAL-Bench/Darwin-60B-DUO | 2026-05-27 | trending=25 | likes=31\nhiiamboris/Darwin-28B-Coder-Q4_K_M-GGUF | 2026-05-21 | trending=5 | likes=5\nk0valik/Darwin-28B-Coder-IQ4_XS12.9GiB-GGUF | 2026-05-21 | trending=5 | likes=5\ntomasmcm/Darwin-4B-Genesis-mlx-4Bit | 2026-05-23 | trending=5 | likes=5\nFlatFootInternational/Darwin-9B-NEG-mlx-fp16 | 2026-05-23 | trending=5 | likes=5\natmzed/Darwin-4B-David-mlx-fp16 | 2026-05-14 | trending=3 | likes=5\natmzed/Darwin-4B-Genesis-mlx-fp16 | 2026-05-14 | trending=3 | likes=5\nFINAL-Bench/Darwin-4B-Genesis | 2026-04-10 | trending=2 | likes=40\nAI-Joe-gi"}]}, "uuid": "0e7a9409-a64d-5a9b-8bab-843f0ab2bfdc", "timestamp": "2026-06-01T12:00:28.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "0e7a9409-a64d-5a9b-8bab-843f0ab2bfdc", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_f6e68fde77e75b5d99acb4", "name": "Bash", "input": {"raw": "{'command': 'hf models info \"Qwen/Qwen3-Coder-30B-A3B-Instruct\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\nd = json.load(sys.stdin)\\nmid = d.get(\\'id\\',\\'\\')\\ncreated = d.get(\\'created_at\\',\\'\\')[:10]\\nparams = d.get(\\'safetensors\\', {}) or {}\\ntotal = params.get(\\'total\\', 0)\\nprint(f\\'{mid} | {created} | {total/1e9:.1f}B | likes={d.get(\\\\\"likes\\\\\")} | trending={d.get(\\\\\"trending_score\\\\\")}\\')\\nprint(\\'tags:\\', [t for t in d.get(\\'tags\\',[]) if \\'param\\' in t.lower() or t.endswith(\\'B\\')])\\n\" 2>/dev/null\\n\\nhf models info \"ibm-granite/granite-3.3-8b-instruct\" --format json 2>"}}]}, "uuid": "9d5867b5-e5da-5586-adf7-7883c13642bf", "timestamp": "2026-06-01T12:00:29.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "9d5867b5-e5da-5586-adf7-7883c13642bf", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_f6e68fde77e75b5d99acb4", "content": "Qwen/Qwen3-Coder-30B-A3B-Instruct | 2025-07-31 | 30.5B | likes=1091 | trending=None\ntags: []\nibm-granite/granite-3.3-8b-instruct | 2025-04-09 | 8.2B | likes=157 | trending=None"}]}, "uuid": "a3f0a090-17b9-5a74-9c41-b19caf5f8350", "timestamp": "2026-06-01T12:00:30.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "a3f0a090-17b9-5a74-9c41-b19caf5f8350", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_6877cc0a8b975689ad6ea9", "name": "Bash", "input": {"raw": "{'command': '# Let me now compile the best candidates. The task asks to \"exclude mainstream Llama, Mistral and Qwen BASE series\"\\n# This means we can include coding-specific models like Qwen3-Coder (it\\'s code-focused fine-tune)\\n# But I need to also check if Qwen3-Coder is considered \"Qwen base series\"\\n# Given the instruction to \"exclude Qwen base series\", Qwen3-Coder is arguably a specialized coding model\\n\\n# Let me check some specific well-known coding models from 2025-2026\\nfor model_id in \"FINAL-Bench/Darwin-28B-Coder\" \"XiaomiMiMo/MiMo-7B-RL\" \"deepseek-ai/DeepSeek-R1-Distill-Llama-8B\" \""}}]}, "uuid": "509c2eb6-10f0-5dcb-9150-44d3a8fbe894", "timestamp": "2026-06-01T12:00:31.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "509c2eb6-10f0-5dcb-9150-44d3a8fbe894", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_6877cc0a8b975689ad6ea9", "content": "Exit code 1\ndeepseek-ai/DeepSeek-R1-Distill-Llama-8B | 2025-01-20 | 8.0B | likes=862 | trending=None"}]}, "uuid": "b06f7aea-3b93-51ab-9caf-e94517e701d0", "timestamp": "2026-06-01T12:00:32.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "b06f7aea-3b93-51ab-9caf-e94517e701d0", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_21eb92218b445a41a566a8", "name": "Bash", "input": {"raw": "{'command': '# Let me look at trending text-generation + coding models more carefully\\n# Also search for specific well-known 2025 coding models \\nhf models list --search \"phi-4\" --filter \"text-generation\" --sort trending_score --limit 10 --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\ndata = json.load(sys.stdin)\\nfor m in data:\\n mid = m.get(\\'id\\',\\'\\')\\n created = m.get(\\'created_at\\',\\'\\')[:10]\\n params = m.get(\\'safetensors\\', {}) or {}\\n total = params.get(\\'total\\', 0)\\n if created >= \\'2025-01-01\\':\\n print(f\\'{mid} | {created} | {total/1e9:.1f}B | tren"}}]}, "uuid": "78a4053f-59f0-5922-a966-7232bd8136f8", "timestamp": "2026-06-01T12:00:33.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "78a4053f-59f0-5922-a966-7232bd8136f8", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_21eb92218b445a41a566a8", "content": "microsoft/Phi-4-mini-instruct | 2025-02-19 | 0.0B | trending=4 | likes=750\nunsloth/Phi-4-reasoning-plus-GGUF | 2025-05-01 | 0.0B | trending=3 | likes=91\nmicrosoft/Phi-4-reasoning | 2025-04-09 | 0.0B | trending=2 | likes=225\nmicrosoft/Phi-4-reasoning-plus | 2025-04-17 | 0.0B | trending=2 | likes=342\nNCUT-AI/Heliars-Phi4-Carla-X1-14B | 2026-05-19 | 0.0B | trending=2 | likes=2\nunsloth/phi-4-GGUF | 2025-01-08 | 0.0B | trending=1 | likes=185\nhuihui-ai/phi-4-abliterated | 2025-01-09 | 0.0B | trending=1 | likes=22\n---\ndeepseek-ai/DeepSeek-R1 | 2025-01-20 | 0.0B | trending=15 | likes=13364\nunsloth/Dee"}]}, "uuid": "9087419d-c2c8-553d-9149-ad68d16a0b35", "timestamp": "2026-06-01T12:00:34.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "9087419d-c2c8-553d-9149-ad68d16a0b35", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_b02b71f5172053e19d32b5", "name": "Bash", "input": {"raw": "{'command': '# Let me now consolidate - I have enough data to compile the top 8 coding models\\n# Key candidates (non-Llama/non-Mistral/non-Qwen base, 2025-2026, <70B params, coding-focused):\\n# 1. deepseek-ai/DeepSeek-R1-Distill-Llama-8B - 8B, code+reasoning, 2025\\n# 2. microsoft/Phi-4-mini-instruct - 3.8B, code capable, 2025\\n# 3. microsoft/Phi-4-reasoning - 14.7B, code+reasoning, 2025\\n# 4. microsoft/Phi-4-reasoning-plus - 14.7B, 2025\\n# 5. JetBrains/Mellum-4b-base - 4B, code-focused, 2025\\n# 6. JetBrains/Mellum2-12B-A2.5B-Thinking - 12.1B, 2026\\n# 7. ByteDance-Seed/Seed-Coder-8B-Base - 8.3B"}}]}, "uuid": "2354d38c-162c-5c6e-8838-2b46bf6249f6", "timestamp": "2026-06-01T12:00:35.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "2354d38c-162c-5c6e-8838-2b46bf6249f6", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_b02b71f5172053e19d32b5", "content": "tags: ['transformers', 'safetensors', 'phi3', 'text-generation', 'nlp', 'code', 'conversational', 'custom_code', 'multilingual', 'ar', 'zh', 'cs', 'da', 'nl', 'en', 'fi', 'fr', 'de', 'he', 'hu']\n---\nByteDance-Seed/Seed-Coder-8B-Base | 2025-04-27 | 0.0B | trending=1 | likes=73\nmlfoundations-dev/DCFT-seed_code_r1_star_coder_instruct-etash | 2025-03-14 | 0.0B | trending=0 | likes=0\nParraga/qwen2.5_32b_coder_it_lora-seed-42 | 2025-04-09 | 0.0B | trending=0 | likes=0\nParraga/qwen2.5_32b_coder_it_lora-seed-666 | 2025-04-09 | 0.0B | trending=0 | likes=0\nByteDance-Seed/Seed-Coder-8B-Instruct | 2025-04"}]}, "uuid": "9265023e-ecc3-5646-aba6-c4acff2d1e0a", "timestamp": "2026-06-01T12:00:36.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "9265023e-ecc3-5646-aba6-c4acff2d1e0a", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_8d7b3ab9916c5aa291dc55", "name": "Bash", "input": {"raw": "{'command': '# Let me also check for Granite code, OpenCoder, and other well-known 2025 coding models\\nfor model_id in \"ibm-granite/granite-3.3-8b-instruct\" \"ibm-granite/granite-3.1-8b-instruct\" \"ibm-granite/granite-34b-code-instruct\" \"infly/OpenCoder-8B-Instruct\" \"ByteDance-Seed/Seed-Coder-8B-Reasoning\"; do\\n result=$(hf models info \"$model_id\" --format json 2>/dev/null)\\n if [ -n \"$result\" ]; then\\n echo \"$result\" | python3 -c \"\\nimport json, sys\\nd = json.load(sys.stdin)\\nmid = d.get(\\'id\\',\\'\\')\\ncreated = d.get(\\'created_at\\',\\'\\')[:10]\\nparams = d.get(\\'safetensors\\', {}) or {}\\ntot"}}]}, "uuid": "f2c5c29d-8a9d-5a26-9c98-e60127249f5a", "timestamp": "2026-06-01T12:00:37.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "f2c5c29d-8a9d-5a26-9c98-e60127249f5a", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_8d7b3ab9916c5aa291dc55", "content": "Exit code 1"}]}, "uuid": "ca850fb5-c85a-54ee-a51c-8566d82542d6", "timestamp": "2026-06-01T12:00:38.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "ca850fb5-c85a-54ee-a51c-8566d82542d6", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_5cfccb76bd4a573c9aa9f6", "name": "Bash", "input": {"raw": "{'command': 'hf models info \"ibm-granite/granite-3.3-8b-instruct\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\nprint(d.get(\\'id\\'), d.get(\\'created_at\\',\\'\\')[:10], d.get(\\'safetensors\\',{}).get(\\'total\\',0)/1e9, \\'likes\\', d.get(\\'likes\\'))\\n\" 2>/dev/null || echo \"not found\"\\n\\nhf models info \"ByteDance-Seed/Seed-Coder-8B-Reasoning\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\nparams = d.get(\\'safetensors\\',{}) or {}\\nprint(d.get(\\'id\\'), d.get(\\'created_at\\',\\'\\')[:10], params.get(\\'total\\',0)/1e9, \\'likes\\', d"}}]}, "uuid": "2c4ced57-9252-5c81-8572-96547b441092", "timestamp": "2026-06-01T12:00:39.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "2c4ced57-9252-5c81-8572-96547b441092", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_5cfccb76bd4a573c9aa9f6", "content": "ibm-granite/granite-3.3-8b-instruct 2025-04-09 8.17086464 likes 157\nByteDance-Seed/Seed-Coder-8B-Reasoning 2025-04-27 8.250462208 likes 154"}]}, "uuid": "ddb9ac91-aef2-512c-a018-b62cffdfdea3", "timestamp": "2026-06-01T12:00:40.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "ddb9ac91-aef2-512c-a018-b62cffdfdea3", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_e92e61308ff058749f65ed", "name": "Bash", "input": {"raw": "{'command': '# Now let me compile a strong final list of 8 trending coding models from 2025-2026, <70B, \\n# excluding Llama/Mistral/Qwen base series\\n# Let me also search using the HF tool for additional candidates\\n\\nhf models list --filter \"text-generation\" --sort trending_score --limit 200 --format json 2>/dev/null | python3 -c \"\\nimport json, sys\\ndata = json.load(sys.stdin)\\nresults = []\\nfor m in data:\\n mid = m.get(\\'id\\',\\'\\')\\n low = mid.lower()\\n # Exclude mainstream base series\\n if any(x in low for x in [\\'llama\\', \\'mistral\\']):\\n continue\\n # Exclude Qwen ba"}}]}, "uuid": "98967aac-09fe-5473-af40-9bcb90394cec", "timestamp": "2026-06-01T12:00:41.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "98967aac-09fe-5473-af40-9bcb90394cec", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_e92e61308ff058749f65ed", "content": "=== CODE-TAGGED MODELS ===\nopenbmb/MiniCPM5-1B | 2026-05-21 | 0B | trending= 347 | likes=741\nstepfun-ai/Step-3.7-Flash | 2026-05-23 | 0B | trending= 218 | likes=221\ndeepseek-ai/DeepSeek-R1 | 2025-01-20 | 0B | trending= 15 | likes=13364\nNemoStation/Marlin-2B | 2026-05-13 | 0B | trending= 105 | likes=504\nstepfun-ai/Step-3.7-Flash-NVFP4 | 2026-05-27 | 0B | trending= 38 | likes=38\nMiniMaxAI/Mi"}]}, "uuid": "f1dbe9bd-6b1d-553a-881f-ee67bb211d08", "timestamp": "2026-06-01T12:00:42.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "f1dbe9bd-6b1d-553a-881f-ee67bb211d08", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_4d3f9f23566b53f7bebfa0", "name": "Bash", "input": {"raw": "{'command': '# Let me get the parameter counts for the top coding candidates\\nfor model_id in \"NemoStation/Marlin-2B\" \"LiquidAI/LFM2.5-8B-A1B\" \"JetBrains/Mellum2-12B-A2.5B-Thinking\" \"JetBrains/Mellum2-12B-A2.5B-Instruct\" \"tencent/Hy-MT2-30B-A3B\" \"tencent/Hy-MT2-1.8B\" \"MiniMaxAI/MiniMax-M2.7\" \"syntropy-ai/Soren-1-Small\" \"poolside/Laguna-XS.2\" \"FINAL-Bench/Darwin-28B-Coder\" \"zai-org/GLM-5.1\"; do\\n result=$(hf models info \"$model_id\" --format json 2>/dev/null)\\n if [ -n \"$result\" ]; then\\n echo \"$result\" | python3 -c \"\\nimport json, sys\\nd = json.load(sys.stdin)\\nparams = d.get(\\'safetensors"}}]}, "uuid": "a2f2b1f1-d571-5413-a779-f5a305e8ddaf", "timestamp": "2026-06-01T12:00:43.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "a2f2b1f1-d571-5413-a779-f5a305e8ddaf", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_4d3f9f23566b53f7bebfa0", "content": "Exit code 1\ntencent/Hy-MT2-1.8B | 2026-05-11 | 2.04B | trending=None | likes=1100 | code=False"}]}, "uuid": "01ae64d3-0c7d-56d7-a1d1-d09c5fb7f0f8", "timestamp": "2026-06-01T12:00:44.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "01ae64d3-0c7d-56d7-a1d1-d09c5fb7f0f8", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_603ee96d37e85464baf141", "name": "Bash", "input": {"raw": "{'command': 'hf models info \"NemoStation/Marlin-2B\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\np = d.get(\\'safetensors\\',{}) or {}\\ntags = \\' \\'.join(d.get(\\'tags\\',[])).lower()\\nprint(d.get(\\'id\\'), d.get(\\'created_at\\',\\'\\')[:10], p.get(\\'total\\',0)/1e9, \\'likes\\', d.get(\\'likes\\'))\\nprint(\\'code tags:\\', [t for t in d.get(\\'tags\\',[]) if \\'code\\' in t.lower() or \\'program\\' in t.lower()])\\n\" 2>/dev/null\\n\\nhf models info \"JetBrains/Mellum2-12B-A2.5B-Thinking\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\np = "}}]}, "uuid": "b66c633d-3c99-5cd7-b4a2-6f56fadfa675", "timestamp": "2026-06-01T12:00:45.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "b66c633d-3c99-5cd7-b4a2-6f56fadfa675", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_603ee96d37e85464baf141", "content": "NemoStation/Marlin-2B 2026-05-13 2.213241664 likes 504\ncode tags: ['custom_code']\nJetBrains/Mellum2-12B-A2.5B-Thinking 2026-05-26 12.149923072 likes 160\ntags: ['transformers', 'safetensors', 'mellum', 'text-generation', 'conversational', 'en', 'arxiv:2605.31268', 'license:apache-2.0', 'model-index', 'eval-results', 'endpoints_compatible', 'region:us']\npoolside/Laguna-XS.2 2026-04-23 33.442617088 likes 284\ntags: ['transformers', 'safetensors', 'laguna', 'text-generation', 'laguna-xs.2', 'vllm', 'conversational', 'custom_code', 'license:apache-2.0', 'eval-results', 'region:us']"}]}, "uuid": "3520cc21-bc57-5e43-bd42-ffaea3754aa1", "timestamp": "2026-06-01T12:00:46.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "3520cc21-bc57-5e43-bd42-ffaea3754aa1", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_9619f4062f78526e9a5092", "name": "Bash", "input": {"raw": "{'command': 'hf models info \"MiniMaxAI/MiniMax-M2.7\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\np = d.get(\\'safetensors\\',{}) or {}\\ntotal = p.get(\\'total\\',0)\\nprint(d.get(\\'id\\'), d.get(\\'created_at\\',\\'\\')[:10], total/1e9, \\'likes\\', d.get(\\'likes\\'))\\nprint(\\'tags:\\', [t for t in d.get(\\'tags\\',[]) if \\'code\\' in t.lower() or \\'math\\' in t.lower()][:10])\\n\" 2>/dev/null\\n\\nhf models info \"zai-org/GLM-5.1\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\np = d.get(\\'safetensors\\',{}) or {}\\ntotal = p.get(\\'total\\"}}]}, "uuid": "b866f17f-e669-5738-990f-a1f0391257d9", "timestamp": "2026-06-01T12:00:47.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "b866f17f-e669-5738-990f-a1f0391257d9", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_9619f4062f78526e9a5092", "content": "MiniMaxAI/MiniMax-M2.7 2026-04-09 228.703644928 likes 1179\ntags: ['custom_code']\nzai-org/GLM-5.1 2026-04-03 753.864139008 likes 1730\ntags: ['transformers', 'safetensors', 'glm_moe_dsa', 'text-generation', 'conversational', 'en', 'zh', 'arxiv:2602.15763', 'license:mit', 'eval-results', 'endpoints_compatible', 'region:us']\ntencent/Hy-MT2-30B-A3B 2026-05-11 30.064725888 likes 445"}]}, "uuid": "49ac7997-ba3a-572d-87df-a846ed7e2535", "timestamp": "2026-06-01T12:00:48.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "49ac7997-ba3a-572d-87df-a846ed7e2535", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_73795028e91050249f6285", "name": "Bash", "input": {"raw": "{'command': '# MiniMax-M2.7 is 228B - too large. GLM-5.1 is 753B - too large. \\n# Let me check more focused coding models\\n\\nhf models info \"LiquidAI/LFM2.5-8B-A1B\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\np = d.get(\\'safetensors\\',{}) or {}\\ntotal = p.get(\\'total\\',0)\\nprint(d.get(\\'id\\'), d.get(\\'created_at\\',\\'\\')[:10], total/1e9, \\'likes\\', d.get(\\'likes\\'))\\nprint(\\'tags:\\', d.get(\\'tags\\',[])[:10])\\n\" 2>/dev/null\\n\\nhf models info \"syntropy-ai/Soren-1-Small\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\n"}}]}, "uuid": "604a8647-3670-56f2-bc28-8dff7ca04c05", "timestamp": "2026-06-01T12:00:49.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "604a8647-3670-56f2-bc28-8dff7ca04c05", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_73795028e91050249f6285", "content": "LiquidAI/LFM2.5-8B-A1B 2026-05-28 8.467856832 likes 457\ntags: ['transformers', 'safetensors', 'lfm2_moe', 'text-generation', 'liquid', 'lfm2.5', 'edge', 'conversational', 'en', 'ar']\nsyntropy-ai/Soren-1-Small 2026-05-20 1.881825088 likes 23\ntags: ['transformers', 'safetensors', 'qwen3_5_text', 'text-generation', 'soren', 'soren-1', 'soren-1-small', 'syntropy-ai', 'project-syntropic', 'chat', 'assistant', 'conversational', 'instruct', 'reasoning', 'thinking']\nGryphe/Pantheon-Reasoning-27B 2026-05-29 27.781427952 likes 20"}]}, "uuid": "b330efe7-ffa4-5c19-8fa5-ccdf21cad608", "timestamp": "2026-06-01T12:00:50.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "b330efe7-ffa4-5c19-8fa5-ccdf21cad608", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_6b5fc5c665255511ab0b64", "name": "Bash", "input": {"command": "# Now let me look for well-known coding models from 2025 - specifically ones designed for coding\n# like Granite Code, DeepSeek-Coder variants from 2025, Phi-4 family for coding\n\n# Mellum2 is JetBrains' coding model - let me check its purpose\nhf models card \"JetBrains/Mellum2-12B-A2.5B-Thinking\" --text 2>/dev/null | head -50\n\necho \"===\"\nhf models card \"JetBrains/Mellum-4b-base\" --text 2>/dev/null | head -30\n"}}]}, "uuid": "ebeb4f31-d7d6-5ea4-9a7d-272f51e5a71c", "timestamp": "2026-06-01T12:00:51.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "ebeb4f31-d7d6-5ea4-9a7d-272f51e5a71c", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_6b5fc5c665255511ab0b64", "content": "<img alt=\"Mellum\" src=\"mellum-logo-dark.svg\" width=\"320\">\n\n# Mellum2 Thinking\n\n> [!Note]\n> Use this model when you want explicit chain-of-thought before the final answer \u2014 complex debugging, multi-step planning, agentic workflows, and math- or reasoning-heavy tasks. For direct, low-latency answers without reasoning traces, use [Instruct](https://huggingface.co/JetBrains/Mellum2-12B-A2.5B-Instruct) instead.\n\n## Mellum2 Thinking Highlights\n\nMellum 2 Thinking is a post-trained reasoning-augmented assistant model trained by JetBrains.\n\nThe model uses a Mixture-of-Experts architecture with 64 exper"}]}, "uuid": "bf04f32f-741e-5cae-95e0-9d0c5d42db9a", "timestamp": "2026-06-01T12:00:52.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "bf04f32f-741e-5cae-95e0-9d0c5d42db9a", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_1e0ea0d58cb25bf69db6e0", "name": "Bash", "input": {"command": "# Great! Now let me get the final confirmed list. Let me also check a few more candidates:\n# - poolside/Laguna-XS.2 (33B, coding-focused)\n# - tencent/Hy-MT2-30B-A3B (30B)\n# - deepseek-ai/DeepSeek-R1-Distill-Llama-8B (8B, reasoning)\n\n# Let me check poolside's Laguna model for coding focus\nhf models card \"poolside/Laguna-XS.2\" --text 2>/dev/null | head -30\n"}}]}, "uuid": "23e0d02b-4973-5a86-998f-21526c29a051", "timestamp": "2026-06-01T12:00:53.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "23e0d02b-4973-5a86-998f-21526c29a051", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_1e0ea0d58cb25bf69db6e0", "content": "<p align=\"center\">\n <img alt=\"poolside-banner\" src=\"https://poolside.ai/assets/laguna/laguna-xs2-banner.svg\" width=\"800px\">\n</p>\n\n<p align=\"center\">\n <a href=\"https://shimmer.poolside.ai\"><strong>Try Laguna XS.2 in Shimmer</strong></a> \u00b7\n <a href=\"https://platform.poolside.ai\"><strong>Get an API key</strong></a> \u00b7\n <a href=\"https://poolside.ai/blog/laguna-a-deeper-dive\"><strong>Release blog post</strong></a>\n</p>\n\n<br>\n\n# Laguna XS.2\nLaguna XS.2 is a 33B total parameter Mixture-of-Experts model with 3B activated parameters per token designed for agentic coding and long-horizon work on a lo"}]}, "uuid": "97183e94-97e6-53d8-bede-2d3a2278ac9c", "timestamp": "2026-06-01T12:00:54.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "97183e94-97e6-53d8-bede-2d3a2278ac9c", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_31eacd9685d651b28ff8a4", "name": "Bash", "input": {"raw": "{'command': '# Excellent! poolside/Laguna-XS.2 is a 33B MoE (3B active) coding model, great fit!\\n# Now let me finalize the 8 models. Let me check a few more\\n\\nhf models info \"deepseek-ai/DeepSeek-R1-Distill-Llama-8B\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\np = d.get(\\'safetensors\\',{}) or {}\\ntotal = p.get(\\'total\\',0)\\nprint(d.get(\\'id\\'), d.get(\\'created_at\\',\\'\\')[:10], f\\'{total/1e9:.2f}B\\', \\'likes\\', d.get(\\'likes\\'))\\nprint(\\'tags:\\', [t for t in d.get(\\'tags\\',[]) if any(x in t.lower() for x in [\\'code\\',\\'reason\\',\\'math\\',\\'think\\'])])\\"}}]}, "uuid": "f0ee173e-177c-5a0b-ae68-29a7dee7d747", "timestamp": "2026-06-01T12:00:55.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "f0ee173e-177c-5a0b-ae68-29a7dee7d747", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_31eacd9685d651b28ff8a4", "content": "deepseek-ai/DeepSeek-R1-Distill-Llama-8B 2025-01-20 8.03B likes 862\ntags: []\nByteDance-Seed/Seed-Coder-8B-Reasoning 2025-04-27 8.25B likes 154\nQwen/Qwen3-Coder-Next 2026-01-30 79.67B likes 1415\nmodel_type: None"}]}, "uuid": "1a255561-24d8-5a82-aa65-a4424a09bee9", "timestamp": "2026-06-01T12:00:56.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "1a255561-24d8-5a82-aa65-a4424a09bee9", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_0da3d4e8a3f9506bbb83e8", "name": "Bash", "input": {"raw": "{'command': '# Qwen3-Coder-Next is 79.67B - exceeds 70B limit! Skip it.\\n# Let me check microsoft/phi-4 (the full phi-4)\\nhf models info \"microsoft/phi-4\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\np = d.get(\\'safetensors\\',{}) or {}\\ntotal = p.get(\\'total\\',0)\\nprint(d.get(\\'id\\'), d.get(\\'created_at\\',\\'\\')[:10], f\\'{total/1e9:.2f}B\\', \\'likes\\', d.get(\\'likes\\'), \\'trending\\', d.get(\\'trending_score\\'))\\n\" 2>/dev/null\\n\\n# Check if Phi-4 is from 2025\\nhf models info \"microsoft/Phi-4-reasoning-plus\" --format json 2>/dev/null | python3 -c \"\\nimport j"}}]}, "uuid": "ff469e36-323e-57a8-8c6a-59a4c652fe00", "timestamp": "2026-06-01T12:00:57.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "ff469e36-323e-57a8-8c6a-59a4c652fe00", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_0da3d4e8a3f9506bbb83e8", "content": "microsoft/phi-4 2024-12-11 14.66B likes 2247 trending None\nmicrosoft/Phi-4-reasoning-plus 2025-04-17 14.66B likes 342 trending None\ntags: ['math', 'code']\nQwen/Qwen3-Coder-30B-A3B-Instruct 2025-07-31 30.53B likes 1091 trending None"}]}, "uuid": "db00bad9-95a6-5d36-9e22-931a011e9c3f", "timestamp": "2026-06-01T12:00:58.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "db00bad9-95a6-5d36-9e22-931a011e9c3f", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_57813be3bfe55925b749f0", "name": "Bash", "input": {"raw": "{'command': '# phi-4 base was 2024 - skip. Phi-4-reasoning-plus is 2025 with math+code tags.\\n# Now let me compile the final 8 trending coding models <70B, 2025-2026, non-Llama/Mistral/Qwen-base\\n\\n# Summary of confirmed candidates:\\n# 1. JetBrains/Mellum2-12B-A2.5B-Thinking - 12.15B, 2026, code-focused (JetBrains), trending=157\\n# 2. JetBrains/Mellum-4b-base - 4.02B, 2025, code-focused (JetBrains), trending=2 (but likes=446)\\n# 3. poolside/Laguna-XS.2 - 33.44B MoE (3B active), 2026, agentic coding, trending=14\\n# 4. microsoft/Phi-4-mini-instruct - 3.84B, 2025, code+multilingual, trending=4\\n#"}}]}, "uuid": "1f0f5fe1-f1fd-5fed-bfb4-6d3a4b2c3228", "timestamp": "2026-06-01T12:00:59.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "1f0f5fe1-f1fd-5fed-bfb4-6d3a4b2c3228", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_57813be3bfe55925b749f0", "content": "# Granite-3.3-8B-Instruct\n\n**Model Summary:**\nGranite-3.3-8B-Instruct is a 8-billion parameter 128K context length language model fine-tuned for improved reasoning and instruction-following capabilities. Built on top of Granite-3.3-8B-Base, the model delivers significant gains on benchmarks for measuring generic performance including AlpacaEval-2.0 and Arena-Hard, and improvements in mathematics, coding, and instruction following. It supports structured reasoning through \\<think\\>\\<\\/think\\> and \\<response\\>\\<\\/response\\> tags, providing clear separation between internal thoughts and final out"}]}, "uuid": "540d7b23-aebf-5258-9bfe-c94e68432c50", "timestamp": "2026-06-01T12:01:00.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "540d7b23-aebf-5258-9bfe-c94e68432c50", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_7c5084bf7f625e3798f7c2", "name": "Bash", "input": {"raw": "{'command': '# Marlin is a video VLM, not a coding model. IBM Granite is more of a general instruction model.\\n# Let me finalize the list with the best 8 coding-focused models\\n\\n# Final confirmed list:\\n# 1. JetBrains/Mellum2-12B-A2.5B-Thinking - 12.15B, 2026, coding (MoE, JetBrains)\\n# 2. JetBrains/Mellum-4b-base - 4.02B, 2025, coding (JetBrains, code completion)\\n# 3. poolside/Laguna-XS.2 - 33.44B (3B active MoE), 2026, agentic coding\\n# 4. microsoft/Phi-4-mini-instruct - 3.84B, 2025, code capable (trending=4)\\n# 5. microsoft/Phi-4-reasoning-plus - 14.66B, 2025, math+code reasoning\\n# 6. By"}}]}, "uuid": "4e323a3d-cb56-5b07-a889-9a939dd22ba8", "timestamp": "2026-06-01T12:01:01.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "4e323a3d-cb56-5b07-a889-9a939dd22ba8", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_7c5084bf7f625e3798f7c2", "content": "===\nnvidia/Qwen3.6-35B-A3B-NVFP4 | 2026-05-27 | 0.0B | trending=136 | likes=143\nnvidia/DeepSeek-V4-Pro-NVFP4 | 2026-05-14 | 0.0B | trending=48 | likes=49\nnvidia/Nemotron-Labs-Diffusion-14B | 2026-04-22 | 0.0B | trending=21 | likes=137\nnvidia/GLM-5.1-NVFP4 | 2026-05-09 | 0.0B | trending=21 | likes=31\nnvidia/Nemotron-Labs-Diffusion-8B | 2026-03-18 | 0.0B | trending=13 | likes=43\nnvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4 | 2026-03-10 | 0.0B | trending=12 | likes=327\nnvidia/Gemma-4-31B-IT-NVFP4 | 2026-04-02 | 0.0B | trending=9 | likes=501\nnvidia/Gemma-4-26B-A4B-NVFP4 | 2026-05-01 | 0.0B | tre"}]}, "uuid": "74745de3-31ca-55fd-afd4-3381f818f77a", "timestamp": "2026-06-01T12:01:02.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "74745de3-31ca-55fd-afd4-3381f818f77a", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_beb9263e9da95abb99e2f1", "name": "Bash", "input": {"raw": "{'command': '# Let me check nvidia/Nemotron-3-Nano and Labs-Diffusion for coding\\nhf models info \"nvidia/Nemotron-Labs-Diffusion-14B\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\np = d.get(\\'safetensors\\',{}) or {}\\ntotal = p.get(\\'total\\',0)\\nprint(d.get(\\'id\\'), d.get(\\'created_at\\',\\'\\')[:10], f\\'{total/1e9:.2f}B\\', \\'likes\\', d.get(\\'likes\\'))\\ntags = d.get(\\'tags\\',[])\\nprint(\\'tags:\\', [t for t in tags if any(x in t.lower() for x in [\\'code\\',\\'coder\\',\\'program\\',\\'diffus\\'])])\\n\" 2>/dev/null\\n\\nhf models info \"nvidia/NVIDIA-Nemotron-3-Nano-30B-A"}}]}, "uuid": "3dca414e-0324-523c-a5cf-93d387fb7e6b", "timestamp": "2026-06-01T12:01:03.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "3dca414e-0324-523c-a5cf-93d387fb7e6b", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_beb9263e9da95abb99e2f1", "content": "nvidia/Nemotron-Labs-Diffusion-14B 2026-04-22 13.51B likes 137\ntags: []\nnvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 2025-12-04 31.58B likes 749\ntags: ['custom_code', 'dataset:nvidia/Nemotron-Pretraining-Code-v1', 'dataset:nvidia/Nemotron-CC-Math-v1', 'dataset:nvidia/Nemotron-Pretraining-Code-v2', 'dataset:nvidia/Nemotron-CC-Code-v1', 'dataset:nvidia/Nemotron-Competitive-Programming-v1', 'dataset:nvidia/Nemotron-Math-v2', 'dataset:nvidia/Nemotron-Math-Proofs-v1']"}]}, "uuid": "bfa5798e-039c-5fda-9e3e-64da96a4ac03", "timestamp": "2026-06-01T12:01:04.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "bfa5798e-039c-5fda-9e3e-64da96a4ac03", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_eb34571ada9c5e09acf1d1", "name": "Bash", "input": {"command": "# NVIDIA Nemotron-3-Nano-30B is trained on code+math datasets! Great fit.\n# Let me read its card to confirm it's code-capable\n\nhf models card \"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16\" --text 2>/dev/null | head -40\n"}}]}, "uuid": "4fd85ae1-fc93-5658-b563-8a50f5243295", "timestamp": "2026-06-01T12:01:05.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "4fd85ae1-fc93-5658-b563-8a50f5243295", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_eb34571ada9c5e09acf1d1", "content": "# NVIDIA-Nemotron-3-Nano-30B-A3B-BF16\n\n<div align=\"center\" style=\"line-height: 1;\">\n<a href=\"https://build.nvidia.com/nvidia/nemotron-3-nano-30b-a3b\" target=\"_blank\" style=\"margin: 2px;\">\n <img alt=\"Chat\" src=\"https://img.shields.io/badge/\ud83e\udd16Chat-Nemotron_3_Nano-536af5?color=76B900&logoColor=white\" style=\"display: inline-block; vertical-align: middle;\"/>\n</a>\n<a href=\"https://arxiv.org/abs/2512.20848\" target=\"_blank\" style=\"margin: 2px;\">\n <img alt=\"Chat\" src=\"https://img.shields.io/badge/\ud83d\udcddPaper-Read Now!-536af5?color=76B900&logoColor=white\" style=\"display: inline-block; vertical-align: mi"}]}, "uuid": "d1f18a6e-e10c-5d64-bdcd-b698cb2668af", "timestamp": "2026-06-01T12:01:06.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "d1f18a6e-e10c-5d64-bdcd-b698cb2668af", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_a18244c8eeb559ea9295b3", "name": "Bash", "input": {"command": "hf models card \"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16\" --text 2>/dev/null | grep -A5 -i \"coding\\|code\\|programming\" | head -30\n"}}]}, "uuid": "2db2602a-0ff5-5247-8737-b09624359b0e", "timestamp": "2026-06-01T12:01:07.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "2db2602a-0ff5-5247-8737-b09624359b0e", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_a18244c8eeb559ea9295b3", "content": "| LiveCodeBench (v6 2025-08\u20132025-05) | **68.3** | 66.0 | 61.0 |\n| SciCode (subtask) | 33.3 | 33.0 | **34.0** |\n| HLE (no tools) | 10.6 | 9.8 | **10.9** |\n| HLE (with tools) | 15.5 | \\- | **17.3** |\n| MiniF2F pass@1 | **50.0** | 5.7 | 12.1 |\n| MiniF2F pass@32 | **79.9** | 16.8 | 43.0 |\n| **Agentic** | | | |\n--\nNVIDIA-Nemotron-3-Nano-30B-A3B-BF16 is a general purpose reasoning and chat model intended to be used in English and coding languages. Other non-English languages (English, Spanish, French, German, Japanese, Italian) are also supported. This model is intended to be used by developers d"}]}, "uuid": "ecc8da52-c3b9-5b9c-94c0-e68d2564832d", "timestamp": "2026-06-01T12:01:08.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "ecc8da52-c3b9-5b9c-94c0-e68d2564832d", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "thinking", "thinking": "Now I have a good list of coding models. Let me compile the final 8:\n\n1. **JetBrains/Mellum2-12B-A2.5B-Thinking** - 12.15B (MoE, 2.5B active), 2026, JetBrains coding model, trending=157\n2. **JetBrains", "signature": ""}]}, "uuid": "cf6c4b62-88d7-57ad-a013-e9ccf5bcc74e", "timestamp": "2026-06-01T12:01:09.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "cf6c4b62-88d7-57ad-a013-e9ccf5bcc74e", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "text", "text": "Now I have enough data to compile the final list. Let me verify one remaining model's parameter count."}]}, "uuid": "c3e021f6-1e70-598b-9894-f491ab14faba", "timestamp": "2026-06-01T12:01:10.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "c3e021f6-1e70-598b-9894-f491ab14faba", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "tool_use", "id": "toolu_2a5fe2abab915792960fa0", "name": "Bash", "input": {"raw": "{'command': 'hf models info \"Qwen/Qwen3-Coder-30B-A3B-Instruct\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\np = d.get(\\'safetensors\\',{}) or {}\\ntotal = p.get(\\'total\\',0)\\nprint(d.get(\\'id\\'), f\\'{total/1e9:.2f}B\\')\\n\" 2>/dev/null\\n\\nhf models info \"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16\" --format json 2>/dev/null | python3 -c \"\\nimport json, sys; d = json.load(sys.stdin)\\np = d.get(\\'safetensors\\',{}) or {}\\ntotal = p.get(\\'total\\',0)\\nprint(d.get(\\'id\\'), f\\'{total/1e9:.2f}B\\')\\n\" 2>/dev/null\\n\\n# Final summary table\\npython3 -c \"\\nmodels = [\\n "}}]}, "uuid": "116b376f-c860-5e7c-9282-daf33cb0202a", "timestamp": "2026-06-01T12:01:11.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "116b376f-c860-5e7c-9282-daf33cb0202a", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "user", "message": {"role": "user", "content": [{"type": "tool_result", "tool_use_id": "toolu_2a5fe2abab915792960fa0", "content": "Qwen/Qwen3-Coder-30B-A3B-Instruct 30.53B\nnvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 31.58B\nJetBrains/Mellum2-12B-A2.5B-Thinking | 12.15B (MoE, 2.5B active) | 2026 | trending=157\nJetBrains/Mellum-4b-base | 4.02B | 2025 | trending=2\npoolside/Laguna-XS.2 | 33.44B (MoE, 3B active) | 2026 | trending=14\nmicrosoft/Phi-4-mini-instruct | 3.84B | 2025 | trending=4\nmicrosoft/Phi-4-reasoning-plus | 14.66B | 2025 | trending=2\nByteDance-Seed/Seed-Coder-8B-Reasoning | 8.25B | 2025 | trending=0\nQwen/Qwen3-Coder-30B-A3B-Instruct | 30.53B (MoE) | 2025 | trending=14\nnvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 | 3"}]}, "uuid": "a612bfdd-79f2-5fb3-91af-f3ba3371af35", "timestamp": "2026-06-01T12:01:12.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
| {"parentUuid": "a612bfdd-79f2-5fb3-91af-f3ba3371af35", "isSidechain": false, "userType": "external", "cwd": "/hub-access-benchmark/claude-code-sonnet/T07/cli_skill", "sessionId": "f86a4801-1626-5e26-8ab9-96b209649034", "version": "1.0.0", "gitBranch": "main", "type": "assistant", "message": {"role": "assistant", "model": "claude-code-sonnet", "content": [{"type": "text", "text": "Here are the 8 trending text-generation models for coding, released in 2025\u20132026, under 70B parameters, suitable for local deployment, excluding mainstream Llama/Mistral/Qwen base series:\n\n| Model ID | Parameters | Released |\n|---|---|---|\n| `JetBrains/Mellum2-12B-A2.5B-Thinking` | 12.15B (MoE, 2.5B active) | 2026 |\n| `poolside/Laguna-XS.2` | 33.44B (MoE, 3B active) | 2026 |\n| `microsoft/Phi-4-mini-instruct` | 3.84B | 2025 |\n| `microsoft/Phi-4-reasoning-plus` | 14.66B | 2025 |\n| `JetBrains/Mellum-4b-base` | 4.02B | 2025 |\n| `ByteDance-Seed/Seed-Coder-8B-Reasoning` | 8.25B | 2025 |\n| `Qwen/Qwen3-Coder-30B-A3B-Instruct` | 30.53B (MoE) | 2025 |\n| `nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16` | 31.58B (MoE, 3B active) | 2025 |\n\n**Notes:**\n- `Qwen3-Coder` is included as it's a coding-specialized fine-tune, not the Qwen base series\n- `Laguna-XS.2` (poolside) is specifically designed for agentic coding / local deployment (runs on Mac 36GB RAM)\n- `Mellum` series are JetBrains' purpose-built code-completion models\n- Several use MoE architecture for efficiency (lower active params than total)\n\nTASK_COMPLETE\n\n```json\n{\n \"models\": [\n {\"id\": \"JetBrains/Mellum2-12B-A2.5B-Thinking\", \"parameters\": \"12.15B (MoE, 2.5B active)\"},\n {\"id\": \"poolside/Laguna-XS.2\", \"parameters\": \"33.44B (MoE, 3B active)\"},\n {\"id\": \"microsoft/Phi-4-mini-instruct\", \"parameters\": \"3.84B\"},\n {\"id\": \"microsoft/Phi-4-reasoning-plus\", \"parameters\": \"14.66B\"},\n {\"id\": \"JetBrains/Mellum-4b-base\", \"parameters\": \"4.02B\"},\n {\"id\": \"ByteDance-Seed/Seed-Coder-8B-Reasoning\", \"parameters\": \"8.25B\"},\n {\"id\": \"Qwen/Qwen3-Coder-30B-A3B-Instruct\", \"parameters\": \"30.53B (MoE)\"},\n {\"id\": \"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16\", \"parameters\": \"31.58B (MoE, 3B active)\"}\n ]\n}\n```"}]}, "uuid": "dab4b60c-69df-5869-b460-9fd07ecea92a", "timestamp": "2026-06-01T12:01:13.000Z", "benchmark_metadata": {"runid": "T07-cli_skill-sonnet-sub-r0-3faf62", "model": "claude-code-sonnet", "task": "T07", "arm": "cli_skill", "rep": 0, "score": 1.0, "verdict": "8 ids not-excl 1.00 code 0.88"}} | |
Xet Storage Details
- Size:
- 91.6 kB
- Xet hash:
- 3f1c28933948fc0d619c8d6fe3b729ed39dc8dcb8a367921b604206b7c8f1807
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.