Duplicate from FINAL-Bench/ALL-Bench-Leaderboard

Co-authored-by: VIDRAFT_LAB <SeaWolf-AI@users.noreply.huggingface.co>

.gitattributes

@@ -0,0 +1,60 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.avro filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.lz4 filter=lfs diff=lfs merge=lfs -text
+*.mds filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+# Audio files - uncompressed
+*.pcm filter=lfs diff=lfs merge=lfs -text
+*.sam filter=lfs diff=lfs merge=lfs -text
+*.raw filter=lfs diff=lfs merge=lfs -text
+# Audio files - compressed
+*.aac filter=lfs diff=lfs merge=lfs -text
+*.flac filter=lfs diff=lfs merge=lfs -text
+*.mp3 filter=lfs diff=lfs merge=lfs -text
+*.ogg filter=lfs diff=lfs merge=lfs -text
+*.wav filter=lfs diff=lfs merge=lfs -text
+# Image files - uncompressed
+*.bmp filter=lfs diff=lfs merge=lfs -text
+*.gif filter=lfs diff=lfs merge=lfs -text
+*.png filter=lfs diff=lfs merge=lfs -text
+*.tiff filter=lfs diff=lfs merge=lfs -text
+# Image files - compressed
+*.jpg filter=lfs diff=lfs merge=lfs -text
+*.jpeg filter=lfs diff=lfs merge=lfs -text
+*.webp filter=lfs diff=lfs merge=lfs -text
+# Video files - compressed
+*.mp4 filter=lfs diff=lfs merge=lfs -text
+*.webm filter=lfs diff=lfs merge=lfs -text

CITATION.cff

@@ -0,0 +1,29 @@
+cff-version: 1.2.0
+title: "ALL Bench Leaderboard 2026: Unified Multi-Modal AI Evaluation"
+message: "If you use this dataset, please cite it as below."
+type: dataset
+authors:
+  - name: "ALL Bench Team"
+url: "https://huggingface.co/spaces/FINAL-Bench/all-bench-leaderboard"
+repository-code: "https://github.com/final-bench/ALL-Bench-Leaderboard"
+license: MIT
+version: "2.1"
+date-released: "2026-03-08"
+keywords:
+  - ai-benchmark
+  - llm-leaderboard
+  - vlm
+  - multimodal-ai
+  - metacognition
+  - final-bench
+  - gpt-5
+  - claude
+  - gemini
+abstract: >-
+  ALL Bench Leaderboard is the only AI benchmark covering LLM, VLM, Agent,
+  Image, Video, and Music generation in a single unified view. It cross-verifies
+  91 AI models across 6 modalities with a 3-tier confidence system. Features
+  composite 5-axis scoring (Knowledge, Expert Reasoning, Abstract Reasoning,
+  Metacognition, Execution), interactive comparison tools, and downloadable
+  intelligence reports. Includes FINAL Bench metacognitive evaluation where
+  Error Recovery explains 94.8% of self-correction performance variance.

README.md

@@ -0,0 +1,333 @@
+---
+---
+annotations_creators:
+  - expert-generated
+language:
+  - en
+license: mit
+pretty_name: "ALL Bench Leaderboard 2026"
+size_categories:
+  - n<1K
+source_datasets:
+  - original
+tags:
+  - benchmark
+  - leaderboard
+  - llm
+  - vlm
+  - ai-evaluation
+  - gpt-5
+  - claude
+  - gemini
+  - final-bench
+  - metacognition
+  - multimodal
+  - ai-agent
+  - image-generation
+  - video-generation
+  - music-generation
+  - union-eval
+task_categories:
+  - text-generation
+  - visual-question-answering
+  - text-to-image
+  - text-to-video
+  - text-to-audio
+configs:
+  - config_name: llm
+    data_files:
+      - split: train
+        path: data/llm.jsonl
+  - config_name: vlm_flagship
+    data_files:
+      - split: train
+        path: data/vlm_flagship.jsonl
+  - config_name: agent
+    data_files:
+      - split: train
+        path: data/agent.jsonl
+  - config_name: image
+    data_files:
+      - split: train
+        path: data/image.jsonl
+  - config_name: video
+    data_files:
+      - split: train
+        path: data/video.jsonl
+  - config_name: music
+    data_files:
+      - split: train
+        path: data/music.jsonl
+models:
+  # LLM - Open Source
+  - Qwen/Qwen3.5-122B-A10B
+  - Qwen/Qwen3.5-27B
+  - Qwen/Qwen3.5-35B-A3B
+  - Qwen/Qwen3.5-9B
+  - Qwen/Qwen3.5-4B
+  - Qwen/Qwen3-Next-80B-A3B-Thinking
+  - deepseek-ai/DeepSeek-V3
+  - deepseek-ai/DeepSeek-R1
+  - zai-org/GLM-5
+  - meta-llama/Llama-4-Scout-17B-16E-Instruct
+  - meta-llama/Llama-4-Maverick-17B-128E-Instruct
+  - microsoft/phi-4
+  - upstage/Solar-Open-100B
+  - K-intelligence/Midm-2.0-Base-Instruct
+  - Nanbeige/Nanbeige4.1-3B
+  - MiniMaxAI/MiniMax-M2.5
+  - stepfun-ai/Step-3.5-Flash
+  # VLM - Open Source
+  - OpenGVLab/InternVL3-78B
+  - Qwen/Qwen2.5-VL-72B-Instruct
+  - Qwen/Qwen3-VL-30B-A3B
+  # Image Generation
+  - black-forest-labs/FLUX.1-dev
+  - stabilityai/stable-diffusion-3.5-large
+  # Video Generation
+  - Lightricks/LTX-Video
+  # Music Generation
+  - facebook/musicgen-large
+  - facebook/jasco-chords-drums-melody-1B
+---
+
+# 🏆 ALL Bench Leaderboard 2026
+
+**The only AI benchmark dataset covering LLM · VLM · Agent · Image · Video · Music in a single unified file.**
+
+<p align="center">
+  <a href="https://huggingface.co/spaces/FINAL-Bench/all-bench-leaderboard"><img src="https://img.shields.io/badge/🏆_Live_Leaderboard-ALL_Bench-6366f1?style=for-the-badge" alt="Live Leaderboard"></a>
+</p>
+
+<p align="center">
+  <a href="https://github.com/final-bench/ALL-Bench-Leaderboard"><img src="https://img.shields.io/badge/GitHub-Repo-black?style=flat-square&logo=github" alt="GitHub"></a>
+  <a href="https://huggingface.co/datasets/FINAL-Bench/Metacognitive"><img src="https://img.shields.io/badge/🧬_FINAL_Bench-Dataset-blueviolet?style=flat-square" alt="FINAL Bench"></a>
+  <a href="https://huggingface.co/spaces/FINAL-Bench/Leaderboard"><img src="https://img.shields.io/badge/🧬_FINAL_Bench-Leaderboard-teal?style=flat-square" alt="FINAL Leaderboard"></a>
+</p>
+
+![ALL Bench Leaderboard](./1.png)
+
+![ALL Bench Leaderboard](./2.png)
+
+
+## Dataset Summary
+
+ALL Bench Leaderboard aggregates and cross-verifies benchmark scores for **90+ AI models** across 6 modalities. Every numerical score is tagged with a confidence level (`cross-verified`, `single-source`, or `self-reported`) and its original source. The dataset is designed for researchers, developers, and decision-makers who need a trustworthy, unified view of the AI model landscape.
+
+| Category | Models | Benchmarks | Description |
+|----------|--------|------------|-------------|
+| **LLM** | 41 | 32 fields | MMLU-Pro, GPQA, AIME, HLE, ARC-AGI-2, Metacog, SWE-Pro, IFEval, LCB, **Union Eval**, etc. |
+| **VLM Flagship** | 11 | 10 fields | MMMU, MMMU-Pro, MathVista, AI2D, OCRBench, MMStar, HallusionBench, etc. |
+| **Agent** | 10 | 8 fields | OSWorld, τ²-bench, BrowseComp, Terminal-Bench 2.0, GDPval-AA, SWE-Pro |
+| **Image Gen** | 10 | 7 fields | Photo realism, text rendering, instruction following, style, aesthetics |
+| **Video Gen** | 10 | 7 fields | Quality, motion, consistency, text rendering, duration, resolution |
+| **Music Gen** | 8 | 6 fields | Quality, vocals, instrumental, lyrics, duration |
+
+![ALL Bench Leaderboard](./3.png)
+
+![ALL Bench Leaderboard](./4.png)
+
+![ALL Bench Leaderboard](./5.png)
+
+
+## What's New — v2.2.1
+
+### 🏅 Union Eval ★NEW
+
+**ALL Bench's proprietary integrated benchmark.** Fuses the discriminative core of 10 existing benchmarks (GPQA, AIME, HLE, MMLU-Pro, IFEval, LiveCodeBench, BFCL, ARC-AGI, SWE, FINAL Bench) into a single 1000-question pool with a season-based rotation system.
+
+**Key features:**
+- **100% JSON auto-graded** — every question requires mandatory JSON output with verifiable fields. Zero keyword matching.
+- **Fuzzy JSON matching** — tolerates key name variants, fraction formats, text fallback when JSON parsing fails.
+- **Season rotation** — 70% new questions each season, 30% anchor questions for cross-season IRT calibration.
+- **8 rounds of empirical testing** — v2 (82.4%) → v3 (82.0%) → Final (79.5%) → S2 (81.8%) → S3 (75.0%) → Fuzzy (69.9/69.3%).
+
+**Key discovery:** *"The bottleneck in benchmarking is not question difficulty — it's grading methodology."*
+
+**Empirically confirmed LLM weakness map:**
+- 🔴 Poetry + code cross-constraints: 18-28%
+- 🔴 Complex JSON structure (10+ constraints): 0%
+- 🔴 Pure series computation (Σk²/3ᵏ): 0%
+- 🟢 Metacognitive reasoning (Bayes, proof errors): 95%
+- 🟢 Revised science detection: 86%
+
+**Current scores (S3, 20Q sample, Fuzzy JSON):**
+
+| Model | Union Eval |
+|-------|-----------|
+| Claude Sonnet 4.6 | **69.9** |
+| Claude Opus 4.6 | **69.3** |
+
+### Other v2.2 changes
+- Fair Coverage Correction: composite scoring ^0.5 → ^0.7
+- +7 FINAL Bench scores (15 total)
+- Columns sorted by fill rate
+- Model Card popup (click model name) · FINAL Bench detail popup (click Metacog score)
+- 🔥 Heatmap, 💰 Price vs Performance scatter tools
+
+
+## Live Leaderboard
+
+👉 **[https://huggingface.co/spaces/FINAL-Bench/all-bench-leaderboard](https://huggingface.co/spaces/FINAL-Bench/all-bench-leaderboard)**
+
+Interactive features: composite ranking, dark mode, advanced search (`GPQA > 90 open`, `price < 1`), Model Finder, Head-to-Head comparison, Trust Map heatmap, Bar Race animation, Model Card popup, FINAL Bench detail popup, and downloadable Intelligence Report (PDF/DOCX).
+
+## Data Structure
+
+```
+data/
+├── llm.jsonl           # 41 LLMs × 32 fields (incl. unionEval ★NEW)
+├── vlm_flagship.jsonl  # 11 flagship VLMs × 10 benchmarks
+├── agent.jsonl         # 10 agent models × 8 benchmarks
+├── image.jsonl         # 10 image gen models × S/A/B/C ratings
+├── video.jsonl         # 10 video gen models × S/A/B/C ratings
+└── music.jsonl         # 8 music gen models × S/A/B/C ratings
+```
+
+## LLM Field Schema
+
+| Field | Type | Description |
+|-------|------|-------------|
+| `name` | string | Model name |
+| `provider` | string | Organization |
+| `type` | string | `open` or `closed` |
+| `group` | string | `flagship`, `open`, `korean`, etc. |
+| `released` | string | Release date (YYYY.MM) |
+| `mmluPro` | float \| null | MMLU-Pro score (%) |
+| `gpqa` | float \| null | GPQA Diamond (%) |
+| `aime` | float \| null | AIME 2025 (%) |
+| `hle` | float \| null | Humanity's Last Exam (%) |
+| `arcAgi2` | float \| null | ARC-AGI-2 (%) |
+| `metacog` | float \| null | FINAL Bench Metacognitive score |
+| `swePro` | float \| null | SWE-bench Pro (%) |
+| `bfcl` | float \| null | Berkeley Function Calling (%) |
+| `ifeval` | float \| null | IFEval instruction following (%) |
+| `lcb` | float \| null | LiveCodeBench (%) |
+| `sweV` | float \| null | SWE-bench Verified (%) — deprecated |
+| `mmmlu` | float \| null | Multilingual MMLU (%) |
+| `termBench` | float \| null | Terminal-Bench 2.0 (%) |
+| `sciCode` | float \| null | SciCode (%) |
+| `unionEval` | float \| null | **★NEW** Union Eval S3 — ALL Bench integrated benchmark (100% JSON auto-graded) |
+| `priceIn` / `priceOut` | float \| null | USD per 1M tokens |
+| `elo` | int \| null | Arena Elo rating |
+| `license` | string | `Prop`, `Apache2`, `MIT`, `Open`, etc. |
+
+![ALL Bench Leaderboard](./6.png)
+
+![ALL Bench Leaderboard](./7.png)
+
+![ALL Bench Leaderboard](./8.png)
+
+## Composite Score
+
+```
+Score = Avg(confirmed benchmarks) × (N/10)^0.7
+```
+
+10 core benchmarks across the **5-Axis Intelligence Framework**: Knowledge · Expert Reasoning · Abstract Reasoning · Metacognition · Execution.
+
+**v2.2 change:** Exponent adjusted from 0.5 to 0.7 for fairer coverage weighting. Models with 7/10 benchmarks receive ×0.79 (was ×0.84), while 4/10 receives ×0.53 (was ×0.63).
+
+## Confidence System
+
+Each benchmark score in the `confidence` object is tagged:
+
+| Level | Badge | Meaning |
+|-------|-------|---------|
+| `cross-verified` | ✓✓ | Confirmed by 2+ independent sources |
+| `single-source` | ✓ | One official or third-party source |
+| `self-reported` | ~ | Provider's own claim, unverified |
+
+Example:
+```json
+"Claude Opus 4.6": {
+  "gpqa": { "level": "cross-verified", "source": "Anthropic + Vellum + DataCamp" },
+  "arcAgi2": { "level": "cross-verified", "source": "Vellum + llm-stats + NxCode + DataCamp" },
+  "metacog": { "level": "single-source", "source": "FINAL Bench dataset" },
+  "unionEval": { "level": "single-source", "source": "Union Eval S3 — ALL Bench official" }
+}
+```
+
+## Usage
+
+```python
+from datasets import load_dataset
+
+# Load LLM data
+ds = load_dataset("FINAL-Bench/ALL-Bench-Leaderboard", "llm")
+df = ds["train"].to_pandas()
+
+# Top 5 LLMs by GPQA
+ranked = df.dropna(subset=["gpqa"]).sort_values("gpqa", ascending=False)
+for _, m in ranked.head(5).iterrows():
+    print(f"{m['name']:25s} GPQA={m['gpqa']}")
+
+# Union Eval scores
+union = df.dropna(subset=["unionEval"]).sort_values("unionEval", ascending=False)
+for _, m in union.iterrows():
+    print(f"{m['name']:25s} Union Eval={m['unionEval']}")
+```
+
+![ALL Bench Leaderboard](./9.png)
+
+![ALL Bench Leaderboard](./10.png)
+
+![ALL Bench Leaderboard](./11.png)
+
+
+## Union Eval — Integrated AI Assessment
+
+Union Eval is ALL Bench's proprietary benchmark designed to address three fundamental problems with existing AI evaluations:
+
+1. **Contamination** — Public benchmarks leak into training data. Union Eval rotates 70% of questions each season.
+2. **Single-axis measurement** — AIME tests only math, IFEval only instruction-following. Union Eval integrates arithmetic, poetry constraints, metacognition, coding, calibration, and myth detection.
+3. **Score inflation via keyword matching** — Traditional rubric grading gives 100% to "well-written" answers even if content is wrong. Union Eval enforces mandatory JSON output with zero keyword matching.
+
+**Structure (S3 — 100 Questions from 1000 Pool):**
+
+| Category | Questions | Role | Expected Score |
+|----------|-----------|------|---------------|
+| Pure Arithmetic | 10 | Confirmed Killer #1 | 0-57% |
+| Poetry/Verse IFEval | 8 | Confirmed Killer #2 | 18-28% |
+| Structured Data IFEval | 7 | JSON/CSV verification | 0-70% |
+| FINAL Bench Metacognition | 20 | Core brand | 50-95% |
+| Union Complex Synthesis | 15 | Extreme multi-domain | 40-73% |
+| Revised Science / Myths | 5 | Calibration traps | 50-86% |
+| Code I/O, GPQA, HLE | 19 | Expert + execution | 50-100% |
+| BFCL Tool Use, Anchors | 16 | Cross-season calibration | varies |
+
+Note: The 100-question dataset is **not publicly released** to prevent contamination. Only scores are published.
+
+
+## FINAL Bench — Metacognitive Benchmark
+
+FINAL Bench measures AI self-correction ability. Error Recovery (ER) explains 94.8% of metacognitive performance variance. 15 frontier models evaluated.
+
+- 🧬 [FINAL-Bench/Metacognitive Dataset](https://huggingface.co/datasets/FINAL-Bench/Metacognitive)
+- 🏆 [FINAL-Bench/Leaderboard](https://huggingface.co/spaces/FINAL-Bench/Leaderboard)
+
+
+## Changelog
+
+| Version | Date | Changes |
+|---------|------|---------|
+| **v2.2.1** | 2026-03-10 | 🏅 **Union Eval ★NEW** — integrated benchmark column (`unionEval` field). Claude Opus 4.6: 69.3 · Sonnet 4.6: 69.9 |
+| v2.2 | 2026-03-10 | Fair Coverage (^0.7), +7 Metacog scores, Model Cards, FINAL Bench popup, Heatmap, Price-Perf |
+| v2.1 | 2026-03-08 | Confidence badges, Intelligence Report, source tracking |
+| v2.0 | 2026-03-07 | All blanks filled, Korean AI data, 42 LLMs cross-verified |
+| v1.9 | 2026-03-05 | +3 LLMs, dark mode, mobile responsive |
+
+## Citation
+
+```bibtex
+@misc{allbench2026,
+    title={ALL Bench Leaderboard 2026: Unified Multi-Modal AI Evaluation},
+    author={ALL Bench Team},
+    year={2026},
+    url={https://huggingface.co/spaces/FINAL-Bench/all-bench-leaderboard}
+}
+```
+
+---
+
+`#AIBenchmark` `#LLMLeaderboard` `#GPT5` `#Claude` `#Gemini` `#ALLBench` `#FINALBench` `#Metacognition` `#UnionEval` `#VLM` `#AIAgent` `#MultiModal` `#HuggingFace` `#ARC-AGI` `#AIEvaluation` `#VIDRAFT.net`

all_bench_leaderboard_v2.1.json

@@ -0,0 +1,3248 @@
+{
+  "metadata": {
+    "name": "ALL Bench Leaderboard 2026",
+    "version": "2.1",
+    "lastVerified": "2026-03-08",
+    "description": "Unified Multi-Modal AI Benchmark — LLM · VLM · Agent · Image · Video · Music",
+    "scoreFormula": "Avg(verified benchmarks) × √(N/10)",
+    "coreBenchmarks": [
+      "MMLU-Pro",
+      "GPQA",
+      "AIME",
+      "HLE",
+      "ARC-AGI-2",
+      "FINAL Bench (Metacog)",
+      "SWE-Pro",
+      "BFCL",
+      "IFEval",
+      "LiveCodeBench"
+    ],
+    "fiveAxisFramework": [
+      "Knowledge",
+      "Expert Reasoning",
+      "Abstract Reasoning",
+      "Metacognition",
+      "Execution"
+    ],
+    "links": {
+      "huggingfaceDataset": "https://huggingface.co/datasets/FINAL-Bench/ALL-Bench-Leaderboard",
+      "github": "https://github.com/final-bench/ALL-Bench-Leaderboard",
+      "finalBenchDataset": "https://huggingface.co/datasets/FINAL-Bench/Metacognitive",
+      "finalBenchLeaderboard": "https://huggingface.co/spaces/FINAL-Bench/Leaderboard"
+    },
+    "counts": {
+      "llm": 42,
+      "vlmFlagship": 11,
+      "vlmLightweight": 5,
+      "agent": 10,
+      "image": 10,
+      "video": 10,
+      "music": 8
+    }
+  },
+  "llm": [
+    {
+      "name": "GPT-5.2",
+      "provider": "OpenAI",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2026.02",
+      "mmluPro": 80,
+      "gpqa": 93.2,
+      "aime": 100,
+      "hle": 35.4,
+      "arcAgi2": 52.9,
+      "metacog": 62.76,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": 90.5,
+      "lcb": 80,
+      "sweV": 80,
+      "mmmlu": 82,
+      "maxIn": 400,
+      "maxOut": 128,
+      "tokPerSec": null,
+      "ttft": 0.6,
+      "vision": [
+        "Image"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning",
+      "elo": 1510,
+      "license": "Prop",
+      "priceIn": 1.75,
+      "priceOut": 14,
+      "modelUrl": "https://openai.com/gpt-5",
+      "termBench": 64.9,
+      "sciCode": 54.6
+    },
+    {
+      "name": "GPT-5.3 Codex",
+      "provider": "OpenAI",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2026.02",
+      "mmluPro": 82.9,
+      "gpqa": 91.5,
+      "aime": 95,
+      "hle": 36,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": 57,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": 83,
+      "maxIn": 400,
+      "maxOut": 128,
+      "tokPerSec": null,
+      "ttft": null,
+      "vision": [
+        "Image"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning(Coding)",
+      "elo": 1500,
+      "license": "Prop",
+      "priceIn": 7.5,
+      "priceOut": 30,
+      "modelUrl": "https://openai.com/codex",
+      "termBench": 77.3,
+      "sciCode": null
+    },
+    {
+      "name": "Claude Opus 4.6",
+      "provider": "Anthropic",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2025.10",
+      "mmluPro": 78.5,
+      "gpqa": 91.3,
+      "aime": 100,
+      "hle": 40,
+      "arcAgi2": 68.8,
+      "metacog": 56.04,
+      "swePro": 45,
+      "bfcl": null,
+      "ifeval": 93.1,
+      "lcb": 76,
+      "sweV": 80.8,
+      "mmmlu": 80,
+      "maxIn": 200,
+      "maxOut": 32,
+      "tokPerSec": null,
+      "ttft": 3.5,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning(Adaptive)",
+      "elo": 1498,
+      "license": "Prop",
+      "priceIn": 5,
+      "priceOut": 25,
+      "modelUrl": "https://anthropic.com/claude",
+      "termBench": 74.7,
+      "sciCode": null
+    },
+    {
+      "name": "Gemini 3.1 Pro",
+      "provider": "Google",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2026.01",
+      "mmluPro": 83,
+      "gpqa": 94.3,
+      "aime": 97,
+      "hle": 44.4,
+      "arcAgi2": 77.1,
+      "metacog": 59.5,
+      "swePro": 43.3,
+      "bfcl": null,
+      "ifeval": 91,
+      "lcb": null,
+      "sweV": 80.6,
+      "mmmlu": 88,
+      "maxIn": 2000,
+      "maxOut": 64,
+      "tokPerSec": null,
+      "ttft": 4.2,
+      "vision": [
+        "Image",
+        "Video",
+        "Audio"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning(DeepThink)",
+      "elo": 1501,
+      "license": "Prop",
+      "priceIn": 2,
+      "priceOut": 12,
+      "modelUrl": "https://gemini.google.com",
+      "termBench": 78.4,
+      "sciCode": 58.9
+    },
+    {
+      "name": "Grok 4 Heavy",
+      "provider": "xAI",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2025.11",
+      "mmluPro": 85,
+      "gpqa": 87.5,
+      "aime": 95,
+      "hle": 28,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": 25,
+      "bfcl": null,
+      "ifeval": 88,
+      "lcb": null,
+      "sweV": 72,
+      "mmmlu": 76,
+      "maxIn": 256,
+      "maxOut": 32,
+      "tokPerSec": null,
+      "ttft": 6,
+      "vision": [
+        "Image"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning",
+      "elo": 1460,
+      "license": "Prop",
+      "priceIn": 3,
+      "priceOut": 15,
+      "modelUrl": "https://x.ai/grok",
+      "termBench": 27.2,
+      "sciCode": null
+    },
+    {
+      "name": "Claude Sonnet 4.6",
+      "provider": "Anthropic",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2026.02",
+      "mmluPro": null,
+      "gpqa": 89.9,
+      "aime": 83,
+      "hle": null,
+      "arcAgi2": 60.4,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": 89.5,
+      "lcb": null,
+      "sweV": 79.6,
+      "mmmlu": null,
+      "maxIn": 200,
+      "maxOut": 64,
+      "tokPerSec": null,
+      "ttft": 2,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning",
+      "elo": 1482,
+      "license": "Prop",
+      "priceIn": 3,
+      "priceOut": 15,
+      "modelUrl": "https://anthropic.com/claude",
+      "termBench": 53,
+      "sciCode": null
+    },
+    {
+      "name": "GPT-OSS-120B",
+      "provider": "OpenAI",
+      "type": "open",
+      "group": "gptoss",
+      "released": "2025.12",
+      "mmluPro": 90,
+      "gpqa": 80.9,
+      "aime": 97.9,
+      "hle": 12,
+      "arcAgi2": null,
+      "metacog": 60.42,
+      "swePro": 16.2,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": 72,
+      "maxIn": 128,
+      "maxOut": 32,
+      "tokPerSec": null,
+      "ttft": null,
+      "vision": [
+        "Text"
+      ],
+      "archType": "MoE",
+      "archDetail": "Reasoning(5.1B/116.8B)",
+      "elo": 1380,
+      "license": "Apache2",
+      "priceIn": 0,
+      "priceOut": 0,
+      "modelUrl": "https://huggingface.co/openai",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "GPT-OSS-20B",
+      "provider": "OpenAI",
+      "type": "open",
+      "group": "gptoss",
+      "released": "2025.12",
+      "mmluPro": 85.3,
+      "gpqa": 71.5,
+      "aime": 98.7,
+      "hle": 7,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": 68,
+      "maxIn": 128,
+      "maxOut": 32,
+      "tokPerSec": null,
+      "ttft": null,
+      "vision": [
+        "Text"
+      ],
+      "archType": "MoE",
+      "archDetail": "Reasoning(3.6B/20.9B)",
+      "elo": 1340,
+      "license": "Apache2",
+      "priceIn": 0,
+      "priceOut": 0,
+      "modelUrl": "https://huggingface.co/openai",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Qwen3.5-397B",
+      "provider": "Alibaba",
+      "type": "open",
+      "group": "qwen",
+      "released": "2026.01",
+      "mmluPro": 87.8,
+      "gpqa": 88.4,
+      "aime": 91.3,
+      "hle": 32,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": 38,
+      "bfcl": null,
+      "ifeval": 92.6,
+      "lcb": null,
+      "sweV": 76.4,
+      "mmmlu": 85,
+      "maxIn": 262,
+      "maxOut": 32,
+      "tokPerSec": 45,
+      "ttft": 5,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "MoE+Hybrid",
+      "archDetail": "Reasoning(17B/397B)",
+      "elo": 1445,
+      "license": "Apache2",
+      "priceIn": null,
+      "priceOut": null,
+      "modelUrl": "https://huggingface.co/Qwen",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Qwen3.5-122B",
+      "provider": "Alibaba",
+      "type": "open",
+      "group": "qwen",
+      "released": "2026.01",
+      "mmluPro": null,
+      "gpqa": 87.6,
+      "aime": 85,
+      "hle": 24,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": 28,
+      "bfcl": 72.2,
+      "ifeval": 93.4,
+      "lcb": null,
+      "sweV": 72.2,
+      "mmmlu": 82,
+      "maxIn": 262,
+      "maxOut": 32,
+      "tokPerSec": null,
+      "ttft": 6,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "MoE+Hybrid",
+      "archDetail": "Reasoning(10B/122B)",
+      "elo": 1420,
+      "license": "Apache2",
+      "priceIn": 0.4,
+      "priceOut": 1.2,
+      "modelUrl": "https://huggingface.co/Qwen",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Qwen3.5-27B",
+      "provider": "Alibaba",
+      "type": "open",
+      "group": "qwen",
+      "released": "2026.01",
+      "mmluPro": 86.1,
+      "gpqa": 85.5,
+      "aime": null,
+      "hle": 18,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": 20,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": 72.4,
+      "mmmlu": 80,
+      "maxIn": 262,
+      "maxOut": 32,
+      "tokPerSec": null,
+      "ttft": 5.5,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning(27B)",
+      "elo": 1395,
+      "license": "Apache2",
+      "priceIn": null,
+      "priceOut": null,
+      "modelUrl": "https://huggingface.co/Qwen",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Qwen3.5-35B",
+      "provider": "Alibaba",
+      "type": "open",
+      "group": "qwen",
+      "released": "2026.01",
+      "mmluPro": null,
+      "gpqa": 83,
+      "aime": null,
+      "hle": 15,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": 18,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": 68,
+      "mmmlu": 78,
+      "maxIn": 262,
+      "maxOut": 32,
+      "tokPerSec": 194,
+      "ttft": 4,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "MoE+Hybrid",
+      "archDetail": "Reasoning(3B/35B)",
+      "elo": 1380,
+      "license": "Apache2",
+      "priceIn": 0.1,
+      "priceOut": 0.4,
+      "modelUrl": "https://huggingface.co/Qwen",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Qwen3.5-Flash",
+      "provider": "Alibaba",
+      "type": "closed",
+      "group": "qwen",
+      "released": "2026.01",
+      "mmluPro": null,
+      "gpqa": null,
+      "aime": null,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": 75,
+      "maxIn": 1000,
+      "maxOut": 32,
+      "tokPerSec": null,
+      "ttft": 3,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "MoE+Hybrid",
+      "archDetail": "Non-Reasoning",
+      "elo": null,
+      "license": "Prop",
+      "priceIn": 0.1,
+      "priceOut": 0.4,
+      "modelUrl": "https://huggingface.co/Qwen",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Qwen3.5-9B",
+      "provider": "Alibaba",
+      "type": "open",
+      "group": "qwen",
+      "released": "2026.01",
+      "mmluPro": 82.5,
+      "gpqa": 81.7,
+      "aime": null,
+      "hle": 10,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": 66.1,
+      "ifeval": 91.5,
+      "lcb": 65.6,
+      "sweV": null,
+      "mmmlu": 81.2,
+      "maxIn": 262,
+      "maxOut": 32,
+      "tokPerSec": null,
+      "ttft": 3.5,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning(9B)",
+      "elo": 1300,
+      "license": "Apache2",
+      "priceIn": null,
+      "priceOut": null,
+      "modelUrl": "https://huggingface.co/Qwen",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Qwen3.5-4B",
+      "provider": "Alibaba",
+      "type": "open",
+      "group": "qwen",
+      "released": "2026.01",
+      "mmluPro": 79.1,
+      "gpqa": 76.2,
+      "aime": null,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": 50.3,
+      "ifeval": 89.8,
+      "lcb": 55.8,
+      "sweV": null,
+      "mmmlu": 76.1,
+      "maxIn": 262,
+      "maxOut": 32,
+      "tokPerSec": null,
+      "ttft": 2,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning(4B)",
+      "elo": null,
+      "license": "Apache2",
+      "priceIn": null,
+      "priceOut": null,
+      "modelUrl": "https://huggingface.co/Qwen",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "DeepSeek V3.2",
+      "provider": "DeepSeek",
+      "type": "open",
+      "group": "value",
+      "released": "2025.12",
+      "mmluPro": null,
+      "gpqa": 79.9,
+      "aime": 89.3,
+      "hle": 22,
+      "arcAgi2": null,
+      "metacog": 60.04,
+      "swePro": 15.6,
+      "bfcl": null,
+      "ifeval": 87.5,
+      "lcb": null,
+      "sweV": 66,
+      "mmmlu": 78,
+      "maxIn": 128,
+      "maxOut": 8,
+      "tokPerSec": null,
+      "ttft": 7,
+      "vision": [
+        "Text"
+      ],
+      "archType": "MoE",
+      "archDetail": "Non-Reasoning(37B/671B)",
+      "elo": 1430,
+      "license": "MIT",
+      "priceIn": 0.14,
+      "priceOut": 0.28,
+      "modelUrl": "https://huggingface.co/deepseek-ai/DeepSeek-V3",
+      "termBench": 39.6,
+      "sciCode": null
+    },
+    {
+      "name": "DeepSeek R1",
+      "provider": "DeepSeek",
+      "type": "open",
+      "group": "value",
+      "released": "2025.01",
+      "mmluPro": 85,
+      "gpqa": 82,
+      "aime": 87.5,
+      "hle": 14,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": 18,
+      "bfcl": null,
+      "ifeval": 83.3,
+      "lcb": null,
+      "sweV": 49.2,
+      "mmmlu": 74,
+      "maxIn": 128,
+      "maxOut": 8,
+      "tokPerSec": null,
+      "ttft": 8,
+      "vision": [
+        "Text"
+      ],
+      "archType": "MoE",
+      "archDetail": "Reasoning(37B/671B)",
+      "elo": 1440,
+      "license": "MIT",
+      "priceIn": 0.55,
+      "priceOut": 2.19,
+      "modelUrl": "https://huggingface.co/deepseek-ai/DeepSeek-R1",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Kimi K2.5",
+      "provider": "Moonshot",
+      "type": "open",
+      "group": "flagship",
+      "released": "2025.11",
+      "mmluPro": 92,
+      "gpqa": 87.6,
+      "aime": 96.1,
+      "hle": 44.9,
+      "arcAgi2": 12.1,
+      "metacog": 68.71,
+      "swePro": 27.7,
+      "bfcl": null,
+      "ifeval": 94,
+      "lcb": 85,
+      "sweV": 76.8,
+      "mmmlu": 81,
+      "maxIn": 200,
+      "maxOut": 32,
+      "tokPerSec": null,
+      "ttft": 5,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "MoE",
+      "archDetail": "Reasoning(1T)",
+      "elo": 1447,
+      "license": "MIT",
+      "priceIn": 0.55,
+      "priceOut": 2.5,
+      "modelUrl": "https://huggingface.co/moonshotai",
+      "termBench": 43.2,
+      "sciCode": null
+    },
+    {
+      "name": "GLM-5",
+      "provider": "Zhipu AI",
+      "type": "open",
+      "group": "value",
+      "released": "2026.02",
+      "mmluPro": 87,
+      "gpqa": 86,
+      "aime": 92.7,
+      "hle": 30.5,
+      "arcAgi2": null,
+      "metacog": 62.5,
+      "swePro": 77.8,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": 52,
+      "sweV": null,
+      "mmmlu": 76,
+      "maxIn": 200,
+      "maxOut": 16,
+      "tokPerSec": null,
+      "ttft": null,
+      "vision": [
+        "Text"
+      ],
+      "archType": "MoE",
+      "archDetail": "Reasoning(40B/745B)",
+      "elo": 1451,
+      "license": "MIT",
+      "priceIn": 0.35,
+      "priceOut": 0.39,
+      "modelUrl": "https://huggingface.co/zai-org/GLM-5",
+      "termBench": 52.4,
+      "sciCode": null
+    },
+    {
+      "name": "Llama 4 Scout",
+      "provider": "Meta",
+      "type": "open",
+      "group": "flagship",
+      "released": "2025.04",
+      "mmluPro": null,
+      "gpqa": 73,
+      "aime": 85,
+      "hle": 12,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": 5.2,
+      "bfcl": null,
+      "ifeval": 85,
+      "lcb": null,
+      "sweV": 55,
+      "mmmlu": 70,
+      "maxIn": 10000,
+      "maxOut": 16,
+      "tokPerSec": null,
+      "ttft": 2,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "MoE",
+      "archDetail": "Non-Reasoning(17B/400B)",
+      "elo": 1340,
+      "license": "Meta",
+      "priceIn": 0.11,
+      "priceOut": 0.34,
+      "modelUrl": "https://huggingface.co/meta-llama",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Mistral Large 3",
+      "provider": "Mistral",
+      "type": "open",
+      "group": "flagship",
+      "released": "2025.11",
+      "mmluPro": null,
+      "gpqa": 78,
+      "aime": 82,
+      "hle": 11,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": 12,
+      "bfcl": null,
+      "ifeval": 86,
+      "lcb": null,
+      "sweV": 60,
+      "mmmlu": 72,
+      "maxIn": 256,
+      "maxOut": 16,
+      "tokPerSec": null,
+      "ttft": 0.3,
+      "vision": [
+        "Image"
+      ],
+      "archType": "MoE",
+      "archDetail": "Non-Reasoning(675B)",
+      "elo": 1320,
+      "license": "Apache2",
+      "priceIn": 2,
+      "priceOut": 6,
+      "modelUrl": "https://huggingface.co/mistralai",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Gemini 3 Flash",
+      "provider": "Google",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2025.12",
+      "mmluPro": 88.6,
+      "gpqa": 90.4,
+      "aime": 95,
+      "hle": 33.7,
+      "arcAgi2": 34,
+      "metacog": null,
+      "swePro": 71.2,
+      "bfcl": null,
+      "ifeval": 88.2,
+      "lcb": null,
+      "sweV": 72.5,
+      "mmmlu": 83,
+      "maxIn": 1000,
+      "maxOut": 64,
+      "tokPerSec": 218,
+      "ttft": 1.2,
+      "vision": [
+        "Image",
+        "Video",
+        "Audio"
+      ],
+      "archType": "Dense",
+      "archDetail": "Non-Reasoning+Thinking",
+      "elo": 1490,
+      "license": "Prop",
+      "priceIn": 0.5,
+      "priceOut": 3,
+      "modelUrl": "https://deepmind.google/technologies/gemini/flash/",
+      "termBench": 64.3,
+      "sciCode": null
+    },
+    {
+      "name": "Llama 4 Maverick",
+      "provider": "Meta",
+      "type": "open",
+      "group": "flagship",
+      "released": "2025.10",
+      "mmluPro": 80.5,
+      "gpqa": 69.8,
+      "aime": 82,
+      "hle": 18,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": 12.3,
+      "bfcl": null,
+      "ifeval": 83,
+      "lcb": null,
+      "sweV": 73,
+      "mmmlu": 74,
+      "maxIn": 1000,
+      "maxOut": 16,
+      "tokPerSec": null,
+      "ttft": 4.5,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "MoE",
+      "archDetail": "Non-Reasoning(17B/400B)",
+      "elo": 1390,
+      "license": "Llama4",
+      "priceIn": 0.22,
+      "priceOut": 0.88,
+      "modelUrl": "https://huggingface.co/meta-llama",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Claude Haiku 4.5",
+      "provider": "Anthropic",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2025.09",
+      "mmluPro": 72,
+      "gpqa": 75,
+      "aime": null,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": 14,
+      "bfcl": null,
+      "ifeval": 86.5,
+      "lcb": null,
+      "sweV": 68,
+      "mmmlu": 71,
+      "maxIn": 200,
+      "maxOut": 8,
+      "tokPerSec": null,
+      "ttft": 0.4,
+      "vision": [
+        "Image"
+      ],
+      "archType": "Dense",
+      "archDetail": "Non-Reasoning",
+      "elo": 1350,
+      "license": "Prop",
+      "priceIn": 1,
+      "priceOut": 5,
+      "modelUrl": "https://anthropic.com/claude",
+      "termBench": 35.5,
+      "sciCode": null
+    },
+    {
+      "name": "Grok 4.1 Fast",
+      "provider": "xAI",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2025.11",
+      "mmluPro": null,
+      "gpqa": 85.3,
+      "aime": 88,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": 72,
+      "maxIn": 2000,
+      "maxOut": 16,
+      "tokPerSec": null,
+      "ttft": 1.5,
+      "vision": [
+        "Image",
+        "Video"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning",
+      "elo": 1380,
+      "license": "Prop",
+      "priceIn": 0.2,
+      "priceOut": 0.8,
+      "modelUrl": "https://x.ai/grok",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "DeepSeek R2",
+      "provider": "DeepSeek",
+      "type": "open",
+      "group": "value",
+      "released": "2026.02",
+      "mmluPro": 87,
+      "gpqa": 88,
+      "aime": 93.8,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": 84,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": 76,
+      "maxIn": 128,
+      "maxOut": 8,
+      "tokPerSec": null,
+      "ttft": 9,
+      "vision": [
+        "Text"
+      ],
+      "archType": "MoE",
+      "archDetail": "Reasoning(671B)",
+      "elo": 1450,
+      "license": "MIT",
+      "priceIn": 0.55,
+      "priceOut": 2.19,
+      "modelUrl": "https://huggingface.co/deepseek-ai",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Phi-4",
+      "provider": "Microsoft",
+      "type": "open",
+      "group": "value",
+      "released": "2024.12",
+      "mmluPro": null,
+      "gpqa": 73,
+      "aime": null,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": 72,
+      "mmmlu": 68,
+      "maxIn": 16,
+      "maxOut": 4,
+      "tokPerSec": null,
+      "ttft": 2.5,
+      "vision": [
+        "Image"
+      ],
+      "archType": "Dense",
+      "archDetail": "Non-Reasoning(14B)",
+      "elo": 1310,
+      "license": "MIT",
+      "priceIn": null,
+      "priceOut": null,
+      "modelUrl": "https://huggingface.co/microsoft/phi-4",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "K-EXAONE",
+      "provider": "LG AI Research",
+      "type": "open",
+      "group": "korean",
+      "released": "2025.12",
+      "mmluPro": 81.8,
+      "gpqa": 75.4,
+      "aime": 85.3,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": 49.4,
+      "mmmlu": 72,
+      "maxIn": 260,
+      "maxOut": 16,
+      "tokPerSec": null,
+      "ttft": null,
+      "vision": [
+        "Image",
+        "Text"
+      ],
+      "archType": "MoE",
+      "archDetail": "Reasoning(236B, K-Foundation)",
+      "elo": null,
+      "license": "Research",
+      "priceIn": 0,
+      "priceOut": 0,
+      "modelUrl": "https://huggingface.co/LGAI-EXAONE",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "A.X K1",
+      "provider": "SK Telecom",
+      "type": "closed",
+      "group": "korean",
+      "released": "2025.12",
+      "mmluPro": null,
+      "gpqa": null,
+      "aime": null,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": 62,
+      "maxIn": 64,
+      "maxOut": 16,
+      "tokPerSec": null,
+      "ttft": null,
+      "vision": [
+        "Text"
+      ],
+      "archType": "MoE",
+      "archDetail": "Reasoning(500B, K-Foundation)",
+      "elo": null,
+      "license": "Apache2",
+      "priceIn": 0,
+      "priceOut": 0,
+      "modelUrl": "https://www.sktelecom.com",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Solar Open 100B",
+      "provider": "Upstage",
+      "type": "open",
+      "group": "korean",
+      "released": "2025.12",
+      "mmluPro": 80.4,
+      "gpqa": 68.1,
+      "aime": 84.3,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": 74.2,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": 68,
+      "maxIn": 100,
+      "maxOut": 16,
+      "tokPerSec": null,
+      "ttft": null,
+      "vision": [
+        "Text"
+      ],
+      "archType": "MoE",
+      "archDetail": "Reasoning(100B, K-Foundation)",
+      "elo": null,
+      "license": "Apache2",
+      "priceIn": 0,
+      "priceOut": 0,
+      "modelUrl": "https://huggingface.co/upstage/Solar-Open-100B",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Motif AI",
+      "provider": "Motif Technologies",
+      "type": "closed",
+      "group": "korean",
+      "released": "2026.02",
+      "mmluPro": null,
+      "gpqa": null,
+      "aime": null,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": 55,
+      "maxIn": 128,
+      "maxOut": 16,
+      "tokPerSec": null,
+      "ttft": null,
+      "vision": [
+        "Text"
+      ],
+      "archType": "Dense",
+      "archDetail": "Non-Reasoning(K-Foundation)",
+      "elo": null,
+      "license": "Prop",
+      "priceIn": 0,
+      "priceOut": 0,
+      "modelUrl": "https://motif.ai",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Mi:dm 2.0 Base",
+      "provider": "KT",
+      "type": "open",
+      "group": "korean",
+      "released": "2025.07",
+      "mmluPro": null,
+      "gpqa": null,
+      "aime": null,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": null,
+      "maxIn": 32,
+      "maxOut": 4,
+      "tokPerSec": null,
+      "ttft": null,
+      "vision": [
+        "Text"
+      ],
+      "archType": "Dense",
+      "archDetail": "Non-Reasoning(11.5B, Llama-DuS)",
+      "elo": null,
+      "license": "MIT",
+      "priceIn": 0,
+      "priceOut": 0,
+      "modelUrl": "https://huggingface.co/K-intelligence/Midm-2.0-Base-Instruct",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Nanbeige4.1-3B",
+      "provider": "Nanbeige",
+      "type": "open",
+      "group": "value",
+      "released": "2026.03",
+      "mmluPro": null,
+      "gpqa": 83.8,
+      "aime": null,
+      "hle": 12.6,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": 56.5,
+      "ifeval": null,
+      "lcb": 76.9,
+      "sweV": null,
+      "mmmlu": null,
+      "maxIn": 131,
+      "maxOut": 131,
+      "tokPerSec": null,
+      "ttft": null,
+      "vision": [
+        "Text"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning(3B)",
+      "elo": null,
+      "license": "Apache2",
+      "priceIn": 0,
+      "priceOut": 0,
+      "modelUrl": "https://huggingface.co/Nanbeige/Nanbeige4.1-3B",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "MiniMax-M2.5",
+      "provider": "MiniMax",
+      "type": "open",
+      "group": "flagship",
+      "released": "2026.02",
+      "mmluPro": 82,
+      "gpqa": 84.8,
+      "aime": 86.3,
+      "hle": 19.1,
+      "arcAgi2": null,
+      "metacog": 60.54,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": 87.5,
+      "lcb": 82.6,
+      "sweV": 80.2,
+      "mmmlu": null,
+      "maxIn": 1000,
+      "maxOut": 64,
+      "tokPerSec": 47,
+      "ttft": 3.3,
+      "vision": [
+        "Text"
+      ],
+      "archType": "MoE",
+      "archDetail": "Reasoning(10B/230B)",
+      "elo": null,
+      "license": "MIT",
+      "priceIn": 0.3,
+      "priceOut": 1.2,
+      "modelUrl": "https://huggingface.co/MiniMaxAI/MiniMax-M2.5",
+      "termBench": 42.2,
+      "sciCode": null
+    },
+    {
+      "name": "Step-3.5-Flash",
+      "provider": "StepFun",
+      "type": "open",
+      "group": "flagship",
+      "released": "2026.02",
+      "mmluPro": null,
+      "gpqa": null,
+      "aime": 97.3,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": 86.4,
+      "sweV": 74.4,
+      "mmmlu": null,
+      "maxIn": 256,
+      "maxOut": 32,
+      "tokPerSec": 300,
+      "ttft": null,
+      "vision": [
+        "Text"
+      ],
+      "archType": "MoE",
+      "archDetail": "Reasoning(11B/196B)",
+      "elo": null,
+      "license": "Apache2",
+      "priceIn": null,
+      "priceOut": null,
+      "modelUrl": "https://huggingface.co/stepfun-ai/Step-3.5-Flash",
+      "termBench": 51,
+      "sciCode": null
+    },
+    {
+      "name": "GPT-5.4",
+      "provider": "OpenAI",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2026.03",
+      "mmluPro": null,
+      "gpqa": 92.8,
+      "aime": null,
+      "hle": 52.1,
+      "arcAgi2": 73.3,
+      "metacog": null,
+      "swePro": 57.7,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": null,
+      "maxIn": 1000,
+      "maxOut": 64,
+      "tokPerSec": null,
+      "ttft": null,
+      "vision": [
+        "Image"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning(Computer Use)",
+      "elo": null,
+      "license": "Prop",
+      "priceIn": 2.5,
+      "priceOut": 15,
+      "modelUrl": "https://openai.com/gpt-5",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "GPT-5.1",
+      "provider": "OpenAI",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2025.11",
+      "mmluPro": null,
+      "gpqa": 88.1,
+      "aime": 94,
+      "hle": 26,
+      "arcAgi2": 17,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": 74.9,
+      "mmmlu": null,
+      "maxIn": 400,
+      "maxOut": 64,
+      "tokPerSec": null,
+      "ttft": 1,
+      "vision": [
+        "Image"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning(Adaptive)",
+      "elo": 1480,
+      "license": "Prop",
+      "priceIn": 1.25,
+      "priceOut": 10,
+      "modelUrl": "https://openai.com/gpt-5",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Gemini 3 Pro",
+      "provider": "Google",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2025.11",
+      "mmluPro": null,
+      "gpqa": 91.9,
+      "aime": 95,
+      "hle": 37.5,
+      "arcAgi2": 31.1,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": 76.2,
+      "mmmlu": 85,
+      "maxIn": 1000,
+      "maxOut": 64,
+      "tokPerSec": 134,
+      "ttft": 3,
+      "vision": [
+        "Image",
+        "Video",
+        "Audio"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning(DeepThink)",
+      "elo": 1490,
+      "license": "Prop",
+      "priceIn": 2,
+      "priceOut": 12,
+      "modelUrl": "https://deepmind.google/technologies/gemini/",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Claude Sonnet 4.5",
+      "provider": "Anthropic",
+      "type": "closed",
+      "group": "flagship",
+      "released": "2025.09",
+      "mmluPro": null,
+      "gpqa": 83.4,
+      "aime": 100,
+      "hle": 30.8,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": 90,
+      "lcb": null,
+      "sweV": 77.2,
+      "mmmlu": null,
+      "maxIn": 200,
+      "maxOut": 64,
+      "tokPerSec": null,
+      "ttft": 1.5,
+      "vision": [
+        "Image"
+      ],
+      "archType": "Dense",
+      "archDetail": "Reasoning(Hybrid)",
+      "elo": 1440,
+      "license": "Prop",
+      "priceIn": 3,
+      "priceOut": 15,
+      "modelUrl": "https://anthropic.com/claude",
+      "termBench": 51,
+      "sciCode": null
+    },
+    {
+      "name": "GPT-5-Nano",
+      "provider": "OpenAI",
+      "type": "closed",
+      "group": "value",
+      "released": "2025.08",
+      "mmluPro": null,
+      "gpqa": null,
+      "aime": null,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": null,
+      "maxIn": 400,
+      "maxOut": 16,
+      "tokPerSec": null,
+      "ttft": 0.2,
+      "vision": [
+        "Image"
+      ],
+      "archType": "Dense",
+      "archDetail": "Non-Reasoning(Nano)",
+      "elo": null,
+      "license": "Prop",
+      "priceIn": 0.05,
+      "priceOut": 0.4,
+      "modelUrl": "https://openai.com/gpt-5",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Gemini 2.5 FL-Lite",
+      "provider": "Google",
+      "type": "closed",
+      "group": "value",
+      "released": "2025.06",
+      "mmluPro": null,
+      "gpqa": null,
+      "aime": null,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": null,
+      "ifeval": null,
+      "lcb": null,
+      "sweV": null,
+      "mmmlu": null,
+      "maxIn": 1000,
+      "maxOut": 64,
+      "tokPerSec": null,
+      "ttft": 0.5,
+      "vision": [
+        "Image",
+        "Video",
+        "Audio"
+      ],
+      "archType": "Dense",
+      "archDetail": "Non-Reasoning+Thinking",
+      "elo": null,
+      "license": "Prop",
+      "priceIn": 0.1,
+      "priceOut": 0.4,
+      "modelUrl": "https://deepmind.google/technologies/gemini/flash/",
+      "termBench": null,
+      "sciCode": null
+    },
+    {
+      "name": "Qwen3-Next-80B",
+      "provider": "Alibaba",
+      "type": "open",
+      "group": "qwen",
+      "released": "2025.09",
+      "mmluPro": 82.7,
+      "gpqa": 77.2,
+      "aime": null,
+      "hle": null,
+      "arcAgi2": null,
+      "metacog": null,
+      "swePro": null,
+      "bfcl": 49.7,
+      "ifeval": 88.9,
+      "lcb": 68.7,
+      "sweV": null,
+      "mmmlu": 81.3,
+      "maxIn": 262,
+      "maxOut": 32,
+      "tokPerSec": null,
+      "ttft": null,
+      "vision": [
+        "Text"
+      ],
+      "archType": "MoE+Hybrid",
+      "archDetail": "Reasoning(3B/80B)",
+      "elo": null,
+      "license": "NCML",
+      "priceIn": 0.15,
+      "priceOut": 1.2,
+      "modelUrl": "https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Thinking",
+      "termBench": null,
+      "sciCode": null
+    }
+  ],
+  "vlm": {
+    "flagship": [
+      {
+        "name": "Gemini 3 Flash",
+        "provider": "Google",
+        "type": "closed",
+        "confidence": "cross-verified",
+        "source": "Vals.ai + Google",
+        "mmmu": 87.6,
+        "mmmuPro": 80,
+        "mathVista": null,
+        "ai2d": null,
+        "ocrBench": null,
+        "mmStar": null,
+        "hallusionBench": null,
+        "mmBenchEN": null,
+        "realWorldQA": null,
+        "videoMME": null
+      },
+      {
+        "name": "Gemini 3 Pro",
+        "provider": "Google",
+        "type": "closed",
+        "confidence": "cross-verified",
+        "source": "Vals.ai + Google DeepMind",
+        "mmmu": 87.5,
+        "mmmuPro": 80,
+        "mathVista": null,
+        "ai2d": null,
+        "ocrBench": null,
+        "mmStar": null,
+        "hallusionBench": null,
+        "mmBenchEN": null,
+        "realWorldQA": null,
+        "videoMME": null
+      },
+      {
+        "name": "Gemini 3.1 Pro",
+        "provider": "Google",
+        "type": "closed",
+        "confidence": "single-source",
+        "source": "Artificial Analysis independent eval",
+        "mmmu": null,
+        "mmmuPro": 82,
+        "mathVista": null,
+        "ai2d": null,
+        "ocrBench": null,
+        "mmStar": null,
+        "hallusionBench": null,
+        "mmBenchEN": null,
+        "realWorldQA": null,
+        "videoMME": null
+      },
+      {
+        "name": "GPT-5.2",
+        "provider": "OpenAI",
+        "type": "closed",
+        "confidence": "cross-verified",
+        "source": "Vals.ai + OpenAI official",
+        "mmmu": 86.7,
+        "mmmuPro": null,
+        "mathVista": null,
+        "ai2d": null,
+        "ocrBench": null,
+        "mmStar": null,
+        "hallusionBench": null,
+        "mmBenchEN": null,
+        "realWorldQA": null,
+        "videoMME": null
+      },
+      {
+        "name": "GPT-5 (original)",
+        "provider": "OpenAI",
+        "type": "closed",
+        "confidence": "single-source",
+        "source": "OpenAI official",
+        "mmmu": 84.2,
+        "mmmuPro": null,
+        "mathVista": null,
+        "ai2d": null,
+        "ocrBench": null,
+        "mmStar": null,
+        "hallusionBench": null,
+        "mmBenchEN": null,
+        "realWorldQA": null,
+        "videoMME": null
+      },
+      {
+        "name": "Claude Opus 4.6",
+        "provider": "Anthropic",
+        "type": "closed",
+        "confidence": "single-source",
+        "source": "Anthropic + digitalapplied",
+        "mmmu": null,
+        "mmmuPro": 85.1,
+        "mathVista": null,
+        "ai2d": null,
+        "ocrBench": null,
+        "mmStar": null,
+        "hallusionBench": null,
+        "mmBenchEN": null,
+        "realWorldQA": null,
+        "videoMME": null
+      },
+      {
+        "name": "Grok 4 Heavy",
+        "provider": "xAI",
+        "type": "closed",
+        "confidence": "single-source",
+        "source": "Vals.ai",
+        "mmmu": 76.5,
+        "mmmuPro": null,
+        "mathVista": null,
+        "ai2d": null,
+        "ocrBench": null,
+        "mmStar": null,
+        "hallusionBench": null,
+        "mmBenchEN": null,
+        "realWorldQA": null,
+        "videoMME": null
+      },
+      {
+        "name": "InternVL3.5-241B",
+        "provider": "OpenGVLab",
+        "type": "open",
+        "confidence": "single-source",
+        "source": "InternVL3.5 paper arXiv",
+        "mmmu": 77.7,
+        "mmmuPro": null,
+        "mathVista": null,
+        "ai2d": null,
+        "ocrBench": null,
+        "mmStar": null,
+        "hallusionBench": null,
+        "mmBenchEN": null,
+        "realWorldQA": null,
+        "videoMME": null
+      },
+      {
+        "name": "InternVL3-78B",
+        "provider": "OpenGVLab",
+        "type": "open",
+        "confidence": "cross-verified",
+        "source": "InternVL3 paper + OpenCompass",
+        "mmmu": 72.2,
+        "mmmuPro": null,
+        "mathVista": 79.6,
+        "ai2d": 89.7,
+        "ocrBench": 90.6,
+        "mmStar": null,
+        "hallusionBench": 59.1,
+        "mmBenchEN": 89,
+        "realWorldQA": 78,
+        "videoMME": null
+      },
+      {
+        "name": "Qwen2.5-VL-72B",
+        "provider": "Alibaba",
+        "type": "open",
+        "confidence": "cross-verified",
+        "source": "DataCamp + Qwen HF model card",
+        "mmmu": 70.2,
+        "mmmuPro": null,
+        "mathVista": 74.8,
+        "ai2d": null,
+        "ocrBench": null,
+        "mmStar": 70.8,
+        "hallusionBench": null,
+        "mmBenchEN": null,
+        "realWorldQA": null,
+        "videoMME": null
+      },
+      {
+        "name": "Kimi-VL-A3B-Thinking",
+        "provider": "Moonshot",
+        "type": "open",
+        "confidence": "single-source",
+        "source": "Moonshot AI official",
+        "mmmu": 64,
+        "mmmuPro": 46.3,
+        "mathVista": 80.1,
+        "ai2d": null,
+        "ocrBench": null,
+        "mmStar": 70.4,
+        "hallusionBench": null,
+        "mmBenchEN": 84.4,
+        "realWorldQA": 70,
+        "videoMME": null
+      }
+    ],
+    "lightweight": [
+      {
+        "name": "GPT-5-Nano",
+        "provider": "OpenAI",
+        "stemPuzzle": {
+          "mmmu": 75.8,
+          "mmmuPro": 57.2,
+          "mathVision": 62.2,
+          "mathVista": 71.5,
+          "weMath": 62.5,
+          "dynaMath": 78,
+          "zeroBench": 1,
+          "zeroBenchSub": 22.2,
+          "vlmsAreBlind": 66.7,
+          "babyVision": 14.4
+        },
+        "generalVQA": {
+          "realWorldQA": 71.8,
+          "mmStar": 68.6,
+          "mmBenchEN": 80.3,
+          "simpleVQA": 46,
+          "hallusionBench": 58.4,
+          "omniDocBench": 55.9,
+          "charXiv": 50.1,
+          "mmLongBench": 31.8,
+          "ccOCR": 58.9,
+          "ai2d": 81.9,
+          "ocrBench": 75.3
+        },
+        "spatialVideoAgentMedical": {
+          "erqa": 45.8,
+          "countBench": 80,
+          "embSpatialBench": 74.2,
+          "refSpatialBench": 12.6,
+          "lingoQA": 57,
+          "videoMMEsub": 71.7,
+          "videoMME": 66.2,
+          "videoMMMU": 63,
+          "mlvu": 69.2,
+          "mmvu": 63.1,
+          "screenSpotPro": null,
+          "osWorld": null,
+          "androidWorld": null,
+          "tirBench": 18.5,
+          "slake": 57,
+          "pmcVQA": 37.8,
+          "medXpertQA": 26.7
+        }
+      },
+      {
+        "name": "Gemini-2.5-FL-Lite",
+        "provider": "Google",
+        "stemPuzzle": {
+          "mmmu": 73.4,
+          "mmmuPro": 59.7,
+          "mathVision": 52.1,
+          "mathVista": 72.8,
+          "weMath": 32.1,
+          "dynaMath": 69.9,
+          "zeroBench": 1,
+          "zeroBenchSub": 19.2,
+          "vlmsAreBlind": 68.4,
+          "babyVision": 17.5
+        },
+        "generalVQA": {
+          "realWorldQA": 72.2,
+          "mmStar": 69.1,
+          "mmBenchEN": 82.7,
+          "simpleVQA": 54.1,
+          "hallusionBench": 64.5,
+          "omniDocBench": 79.4,
+          "charXiv": 56.1,
+          "mmLongBench": 46.5,
+          "ccOCR": 72.9,
+          "ai2d": 85.7,
+          "ocrBench": 82.5
+        },
+        "spatialVideoAgentMedical": {
+          "erqa": 44.3,
+          "countBench": 79.2,
+          "embSpatialBench": 66.1,
+          "refSpatialBench": 11.2,
+          "lingoQA": 17.8,
+          "videoMMEsub": 74.6,
+          "videoMME": 72.7,
+          "videoMMMU": 69.2,
+          "mlvu": 78.5,
+          "mmvu": 65.3,
+          "screenSpotPro": null,
+          "osWorld": null,
+          "androidWorld": null,
+          "tirBench": 21.5,
+          "slake": 65,
+          "pmcVQA": 48.8,
+          "medXpertQA": 35.3
+        }
+      },
+      {
+        "name": "Qwen3-VL-30B-A3B",
+        "provider": "Alibaba",
+        "stemPuzzle": {
+          "mmmu": 76,
+          "mmmuPro": 63,
+          "mathVision": 65.7,
+          "mathVista": 81.9,
+          "weMath": 70,
+          "dynaMath": 80.1,
+          "zeroBench": 0,
+          "zeroBenchSub": 23.7,
+          "vlmsAreBlind": 72.5,
+          "babyVision": 18.6
+        },
+        "generalVQA": {
+          "realWorldQA": 77.4,
+          "mmStar": 75.5,
+          "mmBenchEN": 88.9,
+          "simpleVQA": 54.3,
+          "hallusionBench": 66,
+          "omniDocBench": 86.8,
+          "charXiv": 56.6,
+          "mmLongBench": 47.4,
+          "ccOCR": 77.8,
+          "ai2d": 86.9,
+          "ocrBench": 83.9
+        },
+        "spatialVideoAgentMedical": {
+          "erqa": 45.3,
+          "countBench": 90,
+          "embSpatialBench": 80.6,
+          "refSpatialBench": 54.2,
+          "lingoQA": 62,
+          "videoMMEsub": 79.9,
+          "videoMME": 73.3,
+          "videoMMMU": 75,
+          "mlvu": 78.9,
+          "mmvu": 66.1,
+          "screenSpotPro": 60.5,
+          "osWorld": 30.6,
+          "androidWorld": 55,
+          "tirBench": 22.5,
+          "slake": 68.8,
+          "pmcVQA": 51.5,
+          "medXpertQA": 35.5
+        }
+      },
+      {
+        "name": "Qwen3.5-9B",
+        "provider": "Alibaba",
+        "stemPuzzle": {
+          "mmmu": 78.4,
+          "mmmuPro": 70.1,
+          "mathVision": 78.9,
+          "mathVista": 85.7,
+          "weMath": 75.2,
+          "dynaMath": 83.6,
+          "zeroBench": 3,
+          "zeroBenchSub": 31.1,
+          "vlmsAreBlind": 93.7,
+          "babyVision": "28.6/25.8"
+        },
+        "generalVQA": {
+          "realWorldQA": 80.3,
+          "mmStar": 79.7,
+          "mmBenchEN": 90.1,
+          "simpleVQA": 51.2,
+          "hallusionBench": 69.3,
+          "omniDocBench": 87.7,
+          "charXiv": 73,
+          "mmLongBench": 57.7,
+          "ccOCR": 79.3,
+          "ai2d": 90.2,
+          "ocrBench": 89.2
+        },
+        "spatialVideoAgentMedical": {
+          "erqa": 55.5,
+          "countBench": 97.2,
+          "embSpatialBench": 83,
+          "refSpatialBench": 58.5,
+          "lingoQA": 80.4,
+          "videoMMEsub": 84.5,
+          "videoMME": 78.4,
+          "videoMMMU": 78.9,
+          "mlvu": 84.4,
+          "mmvu": 67.8,
+          "screenSpotPro": 65.2,
+          "osWorld": 41.8,
+          "androidWorld": 57.8,
+          "tirBench": "45.6/31.9",
+          "slake": 79,
+          "pmcVQA": 57.9,
+          "medXpertQA": 49.9
+        }
+      },
+      {
+        "name": "Qwen3.5-4B",
+        "provider": "Alibaba",
+        "stemPuzzle": {
+          "mmmu": 77.6,
+          "mmmuPro": 66.3,
+          "mathVision": 74.6,
+          "mathVista": 85.1,
+          "weMath": 75.4,
+          "dynaMath": 83.3,
+          "zeroBench": 3,
+          "zeroBenchSub": 26.3,
+          "vlmsAreBlind": 92.6,
+          "babyVision": "16.0/19.1"
+        },
+        "generalVQA": {
+          "realWorldQA": 79.5,
+          "mmStar": 78.3,
+          "mmBenchEN": 89.4,
+          "simpleVQA": 43.4,
+          "hallusionBench": 65,
+          "omniDocBench": 86.2,
+          "charXiv": 70.8,
+          "mmLongBench": 54.2,
+          "ccOCR": 76.7,
+          "ai2d": 89.6,
+          "ocrBench": 85
+        },
+        "spatialVideoAgentMedical": {
+          "erqa": 54,
+          "countBench": 96.3,
+          "embSpatialBench": 81.3,
+          "refSpatialBench": 54.6,
+          "lingoQA": 74.4,
+          "videoMMEsub": 83.5,
+          "videoMME": 76.9,
+          "videoMMMU": 74.1,
+          "mlvu": 82.8,
+          "mmvu": 64.9,
+          "screenSpotPro": 60.3,
+          "osWorld": 35.6,
+          "androidWorld": 58.6,
+          "tirBench": "38.9/29.9",
+          "slake": 76.1,
+          "pmcVQA": 55.5,
+          "medXpertQA": 42.9
+        }
+      }
+    ]
+  },
+  "agent": [
+    {
+      "name": "GPT-5.4",
+      "provider": "OpenAI",
+      "osWorld": 75,
+      "tau2Bench": null,
+      "browseComp": 82.7,
+      "termBench2": null,
+      "gdpvalAA": 83,
+      "swePro": null,
+      "screenSpotPro": null,
+      "androidWorld": null
+    },
+    {
+      "name": "Claude Opus 4.6",
+      "provider": "Anthropic",
+      "osWorld": 72.7,
+      "tau2Bench": 91.9,
+      "browseComp": 84,
+      "termBench2": 74.7,
+      "gdpvalAA": 1606,
+      "swePro": 45,
+      "screenSpotPro": null,
+      "androidWorld": null
+    },
+    {
+      "name": "Claude Sonnet 4.6",
+      "provider": "Anthropic",
+      "osWorld": 72.5,
+      "tau2Bench": null,
+      "browseComp": null,
+      "termBench2": 53,
+      "gdpvalAA": 1633,
+      "swePro": null,
+      "screenSpotPro": null,
+      "androidWorld": null
+    },
+    {
+      "name": "Gemini 3.1 Pro",
+      "provider": "Google",
+      "osWorld": null,
+      "tau2Bench": 99.3,
+      "browseComp": 85.9,
+      "termBench2": 78.4,
+      "gdpvalAA": 1317,
+      "swePro": null,
+      "screenSpotPro": null,
+      "androidWorld": null
+    },
+    {
+      "name": "GPT-5.2",
+      "provider": "OpenAI",
+      "osWorld": 38.2,
+      "tau2Bench": 82,
+      "browseComp": 77.9,
+      "termBench2": 64.9,
+      "gdpvalAA": null,
+      "swePro": null,
+      "screenSpotPro": null,
+      "androidWorld": null
+    },
+    {
+      "name": "GPT-5.3 Codex",
+      "provider": "OpenAI",
+      "osWorld": null,
+      "tau2Bench": null,
+      "browseComp": null,
+      "termBench2": 77.3,
+      "gdpvalAA": null,
+      "swePro": 57,
+      "screenSpotPro": null,
+      "androidWorld": null
+    },
+    {
+      "name": "Gemini 3 Flash",
+      "provider": "Google",
+      "osWorld": null,
+      "tau2Bench": null,
+      "browseComp": null,
+      "termBench2": 64.3,
+      "gdpvalAA": null,
+      "swePro": null,
+      "screenSpotPro": null,
+      "androidWorld": null
+    },
+    {
+      "name": "Qwen3.5-9B",
+      "provider": "Alibaba",
+      "osWorld": null,
+      "tau2Bench": 79.9,
+      "browseComp": null,
+      "termBench2": null,
+      "gdpvalAA": null,
+      "swePro": null,
+      "screenSpotPro": 66.1,
+      "androidWorld": 57.8
+    },
+    {
+      "name": "Qwen3.5-4B",
+      "provider": "Alibaba",
+      "osWorld": null,
+      "tau2Bench": 79.1,
+      "browseComp": null,
+      "termBench2": null,
+      "gdpvalAA": null,
+      "swePro": null,
+      "screenSpotPro": 50.3,
+      "androidWorld": 58.6
+    },
+    {
+      "name": "MiniMax-M2.5",
+      "provider": "MiniMax",
+      "osWorld": null,
+      "tau2Bench": null,
+      "browseComp": null,
+      "termBench2": 42.2,
+      "gdpvalAA": null,
+      "swePro": null,
+      "screenSpotPro": null,
+      "androidWorld": null
+    }
+  ],
+  "image": [
+    {
+      "name": "GPT Image 1.5",
+      "provider": "OpenAI",
+      "released": "2026.01",
+      "ranking": "#1",
+      "photoRealism": "S",
+      "textRendering": "A",
+      "instructionFollowing": "S",
+      "style": "S",
+      "aesthetics": "B",
+      "cost": "$$",
+      "license": "Prop"
+    },
+    {
+      "name": "Imagen 4",
+      "provider": "Google",
+      "released": "2025.12",
+      "ranking": "#2",
+      "photoRealism": "S",
+      "textRendering": "A",
+      "instructionFollowing": "S",
+      "style": "A",
+      "aesthetics": "S",
+      "cost": "$$",
+      "license": "Prop"
+    },
+    {
+      "name": "Flux 2 Pro",
+      "provider": "BFL",
+      "released": "2026.01",
+      "ranking": "#3",
+      "photoRealism": "S",
+      "textRendering": "A",
+      "instructionFollowing": "A",
+      "style": "S",
+      "aesthetics": "A",
+      "cost": "$$",
+      "license": "Prop"
+    },
+    {
+      "name": "Midjourney v7",
+      "provider": "Midjourney",
+      "released": "2026.01",
+      "ranking": "#4",
+      "photoRealism": "A",
+      "textRendering": "S",
+      "instructionFollowing": "B",
+      "style": "B",
+      "aesthetics": "A",
+      "cost": "$$",
+      "license": "Prop"
+    },
+    {
+      "name": "Flux 2 Dev",
+      "provider": "BFL",
+      "released": "2026.01",
+      "ranking": "—",
+      "photoRealism": "A",
+      "textRendering": "A",
+      "instructionFollowing": "A",
+      "style": "A",
+      "aesthetics": "A",
+      "cost": "Free",
+      "license": "Apache2"
+    },
+    {
+      "name": "Ideogram 3.0",
+      "provider": "Ideogram",
+      "released": "2025.12",
+      "ranking": "#5",
+      "photoRealism": "A",
+      "textRendering": "B",
+      "instructionFollowing": "S",
+      "style": "A",
+      "aesthetics": "A",
+      "cost": "$",
+      "license": "Prop"
+    },
+    {
+      "name": "DALL-E 3.5",
+      "provider": "OpenAI",
+      "released": "2025.08",
+      "ranking": "#6",
+      "photoRealism": "A",
+      "textRendering": "B",
+      "instructionFollowing": "A",
+      "style": "S",
+      "aesthetics": "A",
+      "cost": "$$",
+      "license": "Prop"
+    },
+    {
+      "name": "Nano Banana 2",
+      "provider": "Google",
+      "released": "2025.12",
+      "ranking": "—",
+      "photoRealism": "A",
+      "textRendering": "A",
+      "instructionFollowing": "B",
+      "style": "A",
+      "aesthetics": "S",
+      "cost": "$",
+      "license": "Prop"
+    },
+    {
+      "name": "SD 3.5",
+      "provider": "Stability AI",
+      "released": "2024.10",
+      "ranking": "—",
+      "photoRealism": "B",
+      "textRendering": "A",
+      "instructionFollowing": "B",
+      "style": "B",
+      "aesthetics": "S",
+      "cost": "Free",
+      "license": "Open"
+    },
+    {
+      "name": "Seedream 4.5",
+      "provider": "ByteDance",
+      "released": "2025.11",
+      "ranking": "—",
+      "photoRealism": "A",
+      "textRendering": "B",
+      "instructionFollowing": "B",
+      "style": "A",
+      "aesthetics": "S",
+      "cost": "$",
+      "license": "Prop"
+    }
+  ],
+  "video": [
+    {
+      "name": "Sora 2",
+      "provider": "OpenAI",
+      "released": "2025.09",
+      "quality": "S",
+      "motion": "S",
+      "consistency": "A",
+      "textRendering": "S",
+      "duration": "20s",
+      "resolution": "1080p",
+      "cost": "$$$",
+      "license": "Prop"
+    },
+    {
+      "name": "Veo 3.1",
+      "provider": "Google",
+      "released": "2026.01",
+      "quality": "S",
+      "motion": "A",
+      "consistency": "S",
+      "textRendering": "S",
+      "duration": "8s",
+      "resolution": "4K",
+      "cost": "$$",
+      "license": "Prop"
+    },
+    {
+      "name": "Runway Gen-4.5",
+      "provider": "Runway",
+      "released": "2026.02",
+      "quality": "S",
+      "motion": "S",
+      "consistency": "A",
+      "textRendering": "A",
+      "duration": "16s",
+      "resolution": "1080p",
+      "cost": "$$",
+      "license": "Prop"
+    },
+    {
+      "name": "Kling 3.0",
+      "provider": "Kuaishou",
+      "released": "2026.02",
+      "quality": "A",
+      "motion": "A",
+      "consistency": "A",
+      "textRendering": "A",
+      "duration": "5min",
+      "resolution": "1080p",
+      "cost": "$",
+      "license": "Prop"
+    },
+    {
+      "name": "Seedance 2.0",
+      "provider": "ByteDance",
+      "released": "2026.01",
+      "quality": "S",
+      "motion": "A",
+      "consistency": "S",
+      "textRendering": "A",
+      "duration": "10s",
+      "resolution": "2K",
+      "cost": "$$",
+      "license": "Prop"
+    },
+    {
+      "name": "Wan 2.6",
+      "provider": "Alibaba",
+      "released": "2026.01",
+      "quality": "A",
+      "motion": "A",
+      "consistency": "B",
+      "textRendering": "A",
+      "duration": "10s",
+      "resolution": "1080p",
+      "cost": "Free",
+      "license": "Open"
+    },
+    {
+      "name": "Pika 2.5",
+      "provider": "Pika",
+      "released": "2025.12",
+      "quality": "B",
+      "motion": "B",
+      "consistency": "B",
+      "textRendering": "A",
+      "duration": "10s",
+      "resolution": "1080p",
+      "cost": "$",
+      "license": "Prop"
+    },
+    {
+      "name": "Luma Ray3",
+      "provider": "Luma",
+      "released": "2026.01",
+      "quality": "A",
+      "motion": "A",
+      "consistency": "B",
+      "textRendering": "B",
+      "duration": "10s",
+      "resolution": "4K",
+      "cost": "$$",
+      "license": "Prop"
+    },
+    {
+      "name": "LTX-2",
+      "provider": "Lightricks",
+      "released": "2026.01",
+      "quality": "A",
+      "motion": "B",
+      "consistency": "A",
+      "textRendering": "B",
+      "duration": "10s",
+      "resolution": "4K",
+      "cost": "Free",
+      "license": "Open"
+    },
+    {
+      "name": "HaiLuo AI",
+      "provider": "MiniMax",
+      "released": "2025.12",
+      "quality": "B",
+      "motion": "A",
+      "consistency": "B",
+      "textRendering": "B",
+      "duration": "6s",
+      "resolution": "1080p",
+      "cost": "$",
+      "license": "Prop"
+    }
+  ],
+  "music": [
+    {
+      "name": "Suno v4.5",
+      "provider": "Suno",
+      "released": "2025.11",
+      "quality": "S",
+      "vocals": "S",
+      "instrumental": "S",
+      "lyrics": "S",
+      "duration": "4min",
+      "cost": "$$",
+      "license": "Prop"
+    },
+    {
+      "name": "Udio v2",
+      "provider": "Udio",
+      "released": "2025.10",
+      "quality": "S",
+      "vocals": "A",
+      "instrumental": "S",
+      "lyrics": "A",
+      "duration": "4min",
+      "cost": "$$",
+      "license": "Prop"
+    },
+    {
+      "name": "Gemini Music",
+      "provider": "Google",
+      "released": "2026.01",
+      "quality": "A",
+      "vocals": "A",
+      "instrumental": "A",
+      "lyrics": "A",
+      "duration": "60s",
+      "cost": "$",
+      "license": "Prop"
+    },
+    {
+      "name": "MusicGen Large",
+      "provider": "Meta",
+      "released": "2024.06",
+      "quality": "B",
+      "vocals": "A",
+      "instrumental": "B",
+      "lyrics": "A",
+      "duration": "30s",
+      "cost": "Free",
+      "license": "Open"
+    },
+    {
+      "name": "Stable Audio 2.0",
+      "provider": "Stability AI",
+      "released": "2024.04",
+      "quality": "B",
+      "vocals": "A",
+      "instrumental": "B",
+      "lyrics": "B",
+      "duration": "3min",
+      "cost": "$",
+      "license": "Open"
+    },
+    {
+      "name": "JASCO",
+      "provider": "Meta",
+      "released": "2024.12",
+      "quality": "C",
+      "vocals": "A",
+      "instrumental": "C",
+      "lyrics": "B",
+      "duration": "30s",
+      "cost": "Free",
+      "license": "Open"
+    },
+    {
+      "name": "Riffusion v2",
+      "provider": "Riffusion",
+      "released": "2025.06",
+      "quality": "B",
+      "vocals": "B",
+      "instrumental": "B",
+      "lyrics": "A",
+      "duration": "30s",
+      "cost": "$",
+      "license": "Prop"
+    },
+    {
+      "name": "Loudme",
+      "provider": "Loudme",
+      "released": "2025.09",
+      "quality": "A",
+      "vocals": "A",
+      "instrumental": "A",
+      "lyrics": "B",
+      "duration": "5min",
+      "cost": "$",
+      "license": "Prop"
+    }
+  ],
+  "confidence": {
+    "GPT-5.4": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "hle": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "swePro": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "lcb": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "arcAgi2": {
+        "level": "cross-verified",
+        "source": "OpenAI + arcprize.org"
+      }
+    },
+    "GPT-5.2": {
+      "mmluPro": {
+        "level": "cross-verified",
+        "source": "OpenAI + Vellum + DataCamp"
+      },
+      "gpqa": {
+        "level": "cross-verified",
+        "source": "OpenAI + Vellum + DataCamp"
+      },
+      "aime": {
+        "level": "cross-verified",
+        "source": "OpenAI + Vellum + DataCamp"
+      },
+      "ifeval": {
+        "level": "cross-verified",
+        "source": "OpenAI + Vellum + DataCamp"
+      },
+      "lcb": {
+        "level": "cross-verified",
+        "source": "OpenAI + Vellum + DataCamp"
+      },
+      "sweV": {
+        "level": "cross-verified",
+        "source": "OpenAI + Vellum + DataCamp"
+      },
+      "hle": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "arcAgi2": {
+        "level": "cross-verified",
+        "source": "OpenAI + arcprize.org"
+      },
+      "mmmlu": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      }
+    },
+    "GPT-5.3 Codex": {
+      "swePro": {
+        "level": "cross-verified",
+        "source": "OpenAI + Scale AI SEAL"
+      },
+      "lcb": {
+        "level": "cross-verified",
+        "source": "OpenAI + Scale AI SEAL"
+      }
+    },
+    "GPT-5.1": {
+      "gpqa": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      }
+    },
+    "GPT-5-Nano": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      }
+    },
+    "GPT-OSS-120B": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      }
+    },
+    "GPT-OSS-20B": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "OpenAI official"
+      }
+    },
+    "Claude Opus 4.6": {
+      "gpqa": {
+        "level": "cross-verified",
+        "source": "Anthropic + Vellum + DataCamp"
+      },
+      "aime": {
+        "level": "cross-verified",
+        "source": "Anthropic + Vellum + NxCode"
+      },
+      "hle": {
+        "level": "cross-verified",
+        "source": "Vellum + digitalapplied"
+      },
+      "arcAgi2": {
+        "level": "cross-verified",
+        "source": "Vellum + llm-stats + NxCode + DataCamp"
+      },
+      "swePro": {
+        "level": "cross-verified",
+        "source": "Anthropic + Scale AI SEAL"
+      },
+      "ifeval": {
+        "level": "cross-verified",
+        "source": "Anthropic + Vellum"
+      },
+      "lcb": {
+        "level": "single-source",
+        "source": "Anthropic official"
+      },
+      "sweV": {
+        "level": "single-source",
+        "source": "Anthropic official"
+      },
+      "metacog": {
+        "level": "single-source",
+        "source": "FINAL Bench dataset"
+      }
+    },
+    "Claude Sonnet 4.6": {
+      "gpqa": {
+        "level": "cross-verified",
+        "source": "Anthropic + Vellum + NxCode"
+      },
+      "aime": {
+        "level": "cross-verified",
+        "source": "Anthropic + Vellum + NxCode"
+      },
+      "arcAgi2": {
+        "level": "cross-verified",
+        "source": "Anthropic + Vellum + NxCode"
+      },
+      "swePro": {
+        "level": "single-source",
+        "source": "Anthropic official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "Anthropic official"
+      },
+      "sweV": {
+        "level": "single-source",
+        "source": "Anthropic official"
+      }
+    },
+    "Claude Sonnet 4.5": {
+      "gpqa": {
+        "level": "cross-verified",
+        "source": "Anthropic + Vellum"
+      },
+      "aime": {
+        "level": "cross-verified",
+        "source": "Anthropic + Vellum"
+      },
+      "swePro": {
+        "level": "single-source",
+        "source": "Anthropic official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "Anthropic official"
+      }
+    },
+    "Claude Haiku 4.5": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Anthropic official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Anthropic official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "Anthropic official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "Anthropic official"
+      }
+    },
+    "Gemini 3.1 Pro": {
+      "gpqa": {
+        "level": "cross-verified",
+        "source": "Google DeepMind + PCMag + NxCode"
+      },
+      "aime": {
+        "level": "cross-verified",
+        "source": "Google DeepMind + Vellum"
+      },
+      "hle": {
+        "level": "cross-verified",
+        "source": "Google DeepMind + Vellum"
+      },
+      "arcAgi2": {
+        "level": "cross-verified",
+        "source": "Google DeepMind + arcprize.org + NxCode + DataCamp"
+      },
+      "sweV": {
+        "level": "cross-verified",
+        "source": "Google DeepMind + marc0.dev"
+      },
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Google DeepMind model card"
+      },
+      "swePro": {
+        "level": "single-source",
+        "source": "Google DeepMind model card"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "Google DeepMind model card"
+      },
+      "lcb": {
+        "level": "single-source",
+        "source": "Google DeepMind model card"
+      },
+      "mmmlu": {
+        "level": "single-source",
+        "source": "Google DeepMind model card"
+      }
+    },
+    "Gemini 3 Pro": {
+      "gpqa": {
+        "level": "single-source",
+        "source": "Google DeepMind"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "Google DeepMind"
+      },
+      "hle": {
+        "level": "single-source",
+        "source": "Google DeepMind"
+      }
+    },
+    "Gemini 3 Flash": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Google DeepMind model card"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Google DeepMind model card"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "Google DeepMind model card"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "Google DeepMind model card"
+      },
+      "lcb": {
+        "level": "single-source",
+        "source": "Google DeepMind model card"
+      },
+      "sweV": {
+        "level": "single-source",
+        "source": "Google DeepMind model card"
+      },
+      "mmmlu": {
+        "level": "single-source",
+        "source": "Google DeepMind model card"
+      }
+    },
+    "Gemini 2.5 FL-Lite": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Google DeepMind"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Google DeepMind"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "Google DeepMind"
+      }
+    },
+    "Grok 4 Heavy": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "xAI official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "xAI official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "xAI official"
+      },
+      "hle": {
+        "level": "single-source",
+        "source": "xAI official"
+      },
+      "arcAgi2": {
+        "level": "single-source",
+        "source": "xAI official"
+      },
+      "swePro": {
+        "level": "single-source",
+        "source": "xAI official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "xAI official"
+      },
+      "lcb": {
+        "level": "single-source",
+        "source": "xAI official"
+      }
+    },
+    "Grok 4.1 Fast": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "xAI official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "xAI official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "xAI official"
+      },
+      "swePro": {
+        "level": "single-source",
+        "source": "xAI official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "xAI official"
+      }
+    },
+    "DeepSeek V3.2": {
+      "mmluPro": {
+        "level": "cross-verified",
+        "source": "DeepSeek + AA Intelligence Index"
+      },
+      "gpqa": {
+        "level": "cross-verified",
+        "source": "DeepSeek + AA Intelligence Index"
+      },
+      "aime": {
+        "level": "cross-verified",
+        "source": "DeepSeek + AA Intelligence Index"
+      },
+      "ifeval": {
+        "level": "cross-verified",
+        "source": "DeepSeek + AA Intelligence Index"
+      },
+      "lcb": {
+        "level": "cross-verified",
+        "source": "DeepSeek + AA Intelligence Index"
+      }
+    },
+    "DeepSeek R1": {
+      "mmluPro": {
+        "level": "cross-verified",
+        "source": "DeepSeek + AA Intelligence Index"
+      },
+      "gpqa": {
+        "level": "cross-verified",
+        "source": "DeepSeek + AA Intelligence Index"
+      },
+      "aime": {
+        "level": "cross-verified",
+        "source": "DeepSeek + AA Intelligence Index"
+      },
+      "ifeval": {
+        "level": "cross-verified",
+        "source": "DeepSeek + AA Intelligence Index"
+      }
+    },
+    "DeepSeek R2": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "DeepSeek official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "DeepSeek official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "DeepSeek official"
+      },
+      "swePro": {
+        "level": "single-source",
+        "source": "DeepSeek official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "DeepSeek official"
+      },
+      "lcb": {
+        "level": "single-source",
+        "source": "DeepSeek official"
+      }
+    },
+    "Kimi K2.5": {
+      "mmluPro": {
+        "level": "cross-verified",
+        "source": "Moonshot AI + AA Intelligence Index"
+      },
+      "gpqa": {
+        "level": "cross-verified",
+        "source": "Moonshot AI + AA Intelligence Index"
+      },
+      "aime": {
+        "level": "cross-verified",
+        "source": "Moonshot AI + AA Intelligence Index"
+      },
+      "hle": {
+        "level": "cross-verified",
+        "source": "Moonshot AI + AA Intelligence Index"
+      },
+      "arcAgi2": {
+        "level": "single-source",
+        "source": "Moonshot AI official"
+      },
+      "metacog": {
+        "level": "single-source",
+        "source": "FINAL Bench dataset"
+      },
+      "lcb": {
+        "level": "single-source",
+        "source": "Moonshot AI official"
+      }
+    },
+    "Qwen3.5-397B": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "swePro": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "lcb": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      }
+    },
+    "Qwen3.5-122B": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      }
+    },
+    "Qwen3.5-27B": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      }
+    },
+    "Qwen3.5-35B": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      }
+    },
+    "Qwen3.5-Flash": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      }
+    },
+    "Qwen3.5-9B": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      }
+    },
+    "Qwen3.5-4B": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      }
+    },
+    "Qwen3-Next-80B": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "Alibaba Cloud official"
+      }
+    },
+    "K-EXAONE": {
+      "mmluPro": {
+        "level": "cross-verified",
+        "source": "Korea Herald + K-EXAONE Technical Report"
+      },
+      "gpqa": {
+        "level": "cross-verified",
+        "source": "Korea Herald + K-EXAONE Technical Report"
+      },
+      "aime": {
+        "level": "cross-verified",
+        "source": "Korea Herald + K-EXAONE Technical Report"
+      },
+      "sweV": {
+        "level": "single-source",
+        "source": "K-EXAONE Technical Report"
+      }
+    },
+    "A.X K1": {
+      "mmluPro": {
+        "level": "self-reported",
+        "source": "Self-reported benchmark"
+      },
+      "gpqa": {
+        "level": "self-reported",
+        "source": "Self-reported benchmark"
+      }
+    },
+    "Solar Open 100B": {
+      "mmluPro": {
+        "level": "self-reported",
+        "source": "Upstage self-reported"
+      },
+      "gpqa": {
+        "level": "self-reported",
+        "source": "Upstage self-reported"
+      }
+    },
+    "Mi:dm 2.0 Base": {
+      "mmluPro": {
+        "level": "self-reported",
+        "source": "Self-reported"
+      }
+    },
+    "GLM-5": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Zhipu AI official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Zhipu AI official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "Zhipu AI official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "Zhipu AI official"
+      },
+      "lcb": {
+        "level": "single-source",
+        "source": "Zhipu AI official"
+      }
+    },
+    "Llama 4 Scout": {
+      "mmluPro": {
+        "level": "cross-verified",
+        "source": "Meta + AA Intelligence Index"
+      },
+      "gpqa": {
+        "level": "cross-verified",
+        "source": "Meta + AA Intelligence Index"
+      },
+      "aime": {
+        "level": "cross-verified",
+        "source": "Meta + AA Intelligence Index"
+      },
+      "ifeval": {
+        "level": "cross-verified",
+        "source": "Meta + AA Intelligence Index"
+      }
+    },
+    "Llama 4 Maverick": {
+      "mmluPro": {
+        "level": "cross-verified",
+        "source": "Meta + AA Intelligence Index"
+      },
+      "gpqa": {
+        "level": "cross-verified",
+        "source": "Meta + AA Intelligence Index"
+      },
+      "aime": {
+        "level": "cross-verified",
+        "source": "Meta + AA Intelligence Index"
+      },
+      "ifeval": {
+        "level": "cross-verified",
+        "source": "Meta + AA Intelligence Index"
+      }
+    },
+    "Mistral Large 3": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Mistral AI official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Mistral AI official"
+      },
+      "ifeval": {
+        "level": "single-source",
+        "source": "Mistral AI official"
+      }
+    },
+    "Phi-4": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "Microsoft official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "Microsoft official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "Microsoft official"
+      }
+    },
+    "MiniMax-M2.5": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "MiniMax official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "MiniMax official"
+      },
+      "aime": {
+        "level": "single-source",
+        "source": "MiniMax official"
+      }
+    },
+    "Step-3.5-Flash": {
+      "mmluPro": {
+        "level": "single-source",
+        "source": "StepFun official"
+      },
+      "gpqa": {
+        "level": "single-source",
+        "source": "StepFun official"
+      }
+    },
+    "Motif AI": {
+      "mmluPro": {
+        "level": "self-reported",
+        "source": "Self-reported"
+      },
+      "gpqa": {
+        "level": "self-reported",
+        "source": "Self-reported"
+      },
+      "aime": {
+        "level": "self-reported",
+        "source": "Self-reported"
+      }
+    },
+    "Nanbeige4.1-3B": {
+      "mmluPro": {
+        "level": "self-reported",
+        "source": "Self-reported"
+      },
+      "gpqa": {
+        "level": "self-reported",
+        "source": "Self-reported"
+      }
+    }
+  }
+}

data/agent.jsonl

@@ -0,0 +1,10 @@
+{"name": "GPT-5.4", "provider": "OpenAI", "osWorld": 75.0, "tau2Bench": null, "browseComp": 82.7, "termBench2": null, "gdpvalAA": 83, "swePro": null, "screenSpotPro": null, "androidWorld": null}
+{"name": "Claude Opus 4.6", "provider": "Anthropic", "osWorld": 72.7, "tau2Bench": 91.9, "browseComp": 84.0, "termBench2": 74.7, "gdpvalAA": 1606, "swePro": 45.0, "screenSpotPro": null, "androidWorld": null}
+{"name": "Claude Sonnet 4.6", "provider": "Anthropic", "osWorld": 72.5, "tau2Bench": null, "browseComp": null, "termBench2": 53.0, "gdpvalAA": 1633, "swePro": null, "screenSpotPro": null, "androidWorld": null}
+{"name": "Gemini 3.1 Pro", "provider": "Google", "osWorld": null, "tau2Bench": 99.3, "browseComp": 85.9, "termBench2": 78.4, "gdpvalAA": 1317, "swePro": null, "screenSpotPro": null, "androidWorld": null}
+{"name": "GPT-5.2", "provider": "OpenAI", "osWorld": 38.2, "tau2Bench": 82.0, "browseComp": 77.9, "termBench2": 64.9, "gdpvalAA": null, "swePro": null, "screenSpotPro": null, "androidWorld": null}
+{"name": "GPT-5.3 Codex", "provider": "OpenAI", "osWorld": null, "tau2Bench": null, "browseComp": null, "termBench2": 77.3, "gdpvalAA": null, "swePro": 57.0, "screenSpotPro": null, "androidWorld": null}
+{"name": "Gemini 3 Flash", "provider": "Google", "osWorld": null, "tau2Bench": null, "browseComp": null, "termBench2": 64.3, "gdpvalAA": null, "swePro": null, "screenSpotPro": null, "androidWorld": null}
+{"name": "Qwen3.5-9B", "provider": "Alibaba", "osWorld": null, "tau2Bench": 79.9, "browseComp": null, "termBench2": null, "gdpvalAA": null, "swePro": null, "screenSpotPro": 66.1, "androidWorld": 57.8}
+{"name": "Qwen3.5-4B", "provider": "Alibaba", "osWorld": null, "tau2Bench": 79.1, "browseComp": null, "termBench2": null, "gdpvalAA": null, "swePro": null, "screenSpotPro": 50.3, "androidWorld": 58.6}
+{"name": "MiniMax-M2.5", "provider": "MiniMax", "osWorld": null, "tau2Bench": null, "browseComp": null, "termBench2": 42.2, "gdpvalAA": null, "swePro": null, "screenSpotPro": null, "androidWorld": null}

data/image.jsonl

@@ -0,0 +1,10 @@
+{"name": "GPT Image 1.5", "provider": "OpenAI", "released": "2026.01", "rank": "#1", "photoRealism": "S", "textRender": "A", "instructFollow": "S", "style": "S", "aesthetics": "B", "price": "$$", "license": "Prop"}
+{"name": "Imagen 4", "provider": "Google", "released": "2025.12", "rank": "#2", "photoRealism": "S", "textRender": "A", "instructFollow": "S", "style": "A", "aesthetics": "S", "price": "$$", "license": "Prop"}
+{"name": "Flux 2 Pro", "provider": "BFL", "released": "2026.01", "rank": "#3", "photoRealism": "S", "textRender": "A", "instructFollow": "A", "style": "S", "aesthetics": "A", "price": "$$", "license": "Prop"}
+{"name": "Midjourney v7", "provider": "Midjourney", "released": "2026.01", "rank": "#4", "photoRealism": "A", "textRender": "S", "instructFollow": "B", "style": "B", "aesthetics": "A", "price": "$$", "license": "Prop"}
+{"name": "Flux 2 Dev", "provider": "BFL", "released": "2026.01", "rank": null, "photoRealism": "A", "textRender": "A", "instructFollow": "A", "style": "A", "aesthetics": "A", "price": "Free", "license": "Apache2"}
+{"name": "Ideogram 3.0", "provider": "Ideogram", "released": "2025.12", "rank": "#5", "photoRealism": "A", "textRender": "B", "instructFollow": "S", "style": "A", "aesthetics": "A", "price": "$", "license": "Prop"}
+{"name": "DALL-E 3.5", "provider": "OpenAI", "released": "2025.08", "rank": "#6", "photoRealism": "A", "textRender": "B", "instructFollow": "A", "style": "S", "aesthetics": "A", "price": "$$", "license": "Prop"}
+{"name": "Nano Banana 2", "provider": "Google", "released": "2025.12", "rank": null, "photoRealism": "A", "textRender": "A", "instructFollow": "B", "style": "A", "aesthetics": "S", "price": "$", "license": "Prop"}
+{"name": "SD 3.5", "provider": "Stability AI", "released": "2024.10", "rank": null, "photoRealism": "B", "textRender": "A", "instructFollow": "B", "style": "B", "aesthetics": "S", "price": "Free", "license": "Open"}
+{"name": "Seedream 4.5", "provider": "ByteDance", "released": "2025.11", "rank": null, "photoRealism": "A", "textRender": "B", "instructFollow": "B", "style": "A", "aesthetics": "S", "price": "$", "license": "Prop"}

data/llm.jsonl

@@ -0,0 +1,41 @@
+{"name": "GPT-5.2", "provider": "OpenAI", "aime": 100, "arcAgi2": 52.9, "arch": "Dense", "archDetail": "Reasoning", "bfcl": null, "ctxIn": 400, "ctxOut": 128, "elo": 1510, "gpqa": 93.2, "group": "flagship", "hle": 35.4, "ifeval": 90.5, "lcb": 80.0, "license": "Prop", "metacog": 62.76, "mmluPro": 80.0, "mmmlu": 82, "priceIn": 1.75, "priceOut": 14.0, "released": "2026.02", "sciCode": 54.6, "swePro": null, "sweV": 80.0, "termBench": 64.9, "tokPerSec": null, "ttft": 0.6, "type": "closed", "unionEval": null, "url": "https://openai.com/gpt-5", "vision": "Image"}
+{"name": "GPT-5.3 Codex", "provider": "OpenAI", "aime": 95.0, "arcAgi2": null, "arch": "Dense", "archDetail": "Reasoning(Coding)", "bfcl": null, "ctxIn": 400, "ctxOut": 128, "elo": 1500, "gpqa": 91.5, "group": "flagship", "hle": 36.0, "ifeval": null, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": 82.9, "mmmlu": 83, "priceIn": 7.5, "priceOut": 30.0, "released": "2026.02", "sciCode": null, "swePro": 57.0, "sweV": null, "termBench": 77.3, "tokPerSec": null, "ttft": null, "type": "closed", "unionEval": null, "url": "https://openai.com/codex", "vision": "Image"}
+{"name": "Claude Opus 4.6", "provider": "Anthropic", "aime": 100, "arcAgi2": 68.8, "arch": "Dense", "archDetail": "Reasoning(Adaptive)", "bfcl": null, "ctxIn": 200, "ctxOut": 32, "elo": 1498, "gpqa": 91.3, "group": "flagship", "hle": 40.0, "ifeval": 93.1, "lcb": 76.0, "license": "Prop", "metacog": 56.04, "mmluPro": 78.5, "mmmlu": 80, "priceIn": 5.0, "priceOut": 25.0, "released": "2025.10", "sciCode": null, "swePro": 45.0, "sweV": 80.8, "termBench": 74.7, "tokPerSec": null, "ttft": 3.5, "type": "closed", "unionEval": 69.3, "url": "https://anthropic.com/claude", "vision": "Image, Video"}
+{"name": "Gemini 3.1 Pro", "provider": "Google", "aime": 97, "arcAgi2": 77.1, "arch": "Dense", "archDetail": "Reasoning(DeepThink)", "bfcl": null, "ctxIn": 2000, "ctxOut": 64, "elo": 1501, "gpqa": 94.3, "group": "flagship", "hle": 44.4, "ifeval": 91.0, "lcb": null, "license": "Prop", "metacog": 59.5, "mmluPro": 83.0, "mmmlu": 88, "priceIn": 2.0, "priceOut": 12.0, "released": "2026.01", "sciCode": 58.9, "swePro": 43.3, "sweV": 80.6, "termBench": 78.4, "tokPerSec": null, "ttft": 4.2, "type": "closed", "unionEval": null, "url": "https://gemini.google.com", "vision": "Image, Video, Audio"}
+{"name": "Grok 4 Heavy", "provider": "xAI", "aime": 95, "arcAgi2": null, "arch": "Dense", "archDetail": "Reasoning", "bfcl": null, "ctxIn": 256, "ctxOut": 32, "elo": 1460, "gpqa": 87.5, "group": "flagship", "hle": 28.0, "ifeval": 88.0, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": 85.0, "mmmlu": 76, "priceIn": 3.0, "priceOut": 15.0, "released": "2025.11", "sciCode": null, "swePro": 25.0, "sweV": 72.0, "termBench": 27.2, "tokPerSec": null, "ttft": 6.0, "type": "closed", "unionEval": null, "url": "https://x.ai/grok", "vision": "Image"}
+{"name": "Claude Sonnet 4.6", "provider": "Anthropic", "aime": 83.0, "arcAgi2": 60.4, "arch": "Dense", "archDetail": "Reasoning", "bfcl": null, "ctxIn": 200, "ctxOut": 64, "elo": 1482, "gpqa": 89.9, "group": "flagship", "hle": null, "ifeval": 89.5, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": null, "mmmlu": null, "priceIn": 3.0, "priceOut": 15.0, "released": "2026.02", "sciCode": null, "swePro": null, "sweV": 79.6, "termBench": 53.0, "tokPerSec": null, "ttft": 2.0, "type": "closed", "unionEval": 69.9, "url": "https://anthropic.com/claude", "vision": "Image, Video"}
+{"name": "GPT-OSS-120B", "provider": "OpenAI", "aime": 97.9, "arcAgi2": null, "arch": "MoE", "archDetail": "Reasoning(5.1B/116.8B)", "bfcl": null, "ctxIn": 128, "ctxOut": 32, "elo": 1380, "gpqa": 80.9, "group": "gptoss", "hle": 12.0, "ifeval": null, "lcb": null, "license": "Apache2", "metacog": 60.42, "mmluPro": 90.0, "mmmlu": 72, "priceIn": 0, "priceOut": 0, "released": "2025.12", "sciCode": null, "swePro": 16.2, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": null, "type": "open", "unionEval": null, "url": "https://huggingface.co/openai", "vision": "Text"}
+{"name": "GPT-OSS-20B", "provider": "OpenAI", "aime": 98.7, "arcAgi2": null, "arch": "MoE", "archDetail": "Reasoning(3.6B/20.9B)", "bfcl": null, "ctxIn": 128, "ctxOut": 32, "elo": 1340, "gpqa": 71.5, "group": "gptoss", "hle": 7.0, "ifeval": null, "lcb": null, "license": "Apache2", "metacog": null, "mmluPro": 85.3, "mmmlu": 68, "priceIn": 0, "priceOut": 0, "released": "2025.12", "sciCode": null, "swePro": null, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": null, "type": "open", "unionEval": null, "url": "https://huggingface.co/openai", "vision": "Text"}
+{"name": "Qwen3.5-397B", "provider": "Alibaba", "aime": 91.3, "arcAgi2": null, "arch": "MoE+Hybrid", "archDetail": "Reasoning(17B/397B)", "bfcl": null, "ctxIn": 262, "ctxOut": 32, "elo": 1445, "gpqa": 88.4, "group": "qwen", "hle": 32.0, "ifeval": 92.6, "lcb": null, "license": "Apache2", "metacog": 60.9, "mmluPro": 87.8, "mmmlu": 85, "priceIn": null, "priceOut": null, "released": "2026.01", "sciCode": null, "swePro": 38.0, "sweV": 76.4, "termBench": null, "tokPerSec": 45, "ttft": 5.0, "type": "open", "unionEval": null, "url": "https://huggingface.co/Qwen", "vision": "Image, Video"}
+{"name": "Qwen3.5-122B", "provider": "Alibaba", "aime": 85.0, "arcAgi2": null, "arch": "MoE+Hybrid", "archDetail": "Reasoning(10B/122B)", "bfcl": 72.2, "ctxIn": 262, "ctxOut": 32, "elo": 1420, "gpqa": 87.6, "group": "qwen", "hle": 24.0, "ifeval": 93.4, "lcb": null, "license": "Apache2", "metacog": 60.2, "mmluPro": null, "mmmlu": 82, "priceIn": 0.4, "priceOut": 1.2, "released": "2026.01", "sciCode": null, "swePro": 28.0, "sweV": 72.2, "termBench": null, "tokPerSec": null, "ttft": 6.0, "type": "open", "unionEval": null, "url": "https://huggingface.co/Qwen", "vision": "Image, Video"}
+{"name": "Qwen3.5-27B", "provider": "Alibaba", "aime": null, "arcAgi2": null, "arch": "Dense", "archDetail": "Reasoning(27B)", "bfcl": null, "ctxIn": 262, "ctxOut": 32, "elo": 1395, "gpqa": 85.5, "group": "qwen", "hle": 18.0, "ifeval": null, "lcb": null, "license": "Apache2", "metacog": 60.6, "mmluPro": 86.1, "mmmlu": 80, "priceIn": null, "priceOut": null, "released": "2026.01", "sciCode": null, "swePro": 20.0, "sweV": 72.4, "termBench": null, "tokPerSec": null, "ttft": 5.5, "type": "open", "unionEval": null, "url": "https://huggingface.co/Qwen", "vision": "Image, Video"}
+{"name": "Qwen3.5-35B", "provider": "Alibaba", "aime": null, "arcAgi2": null, "arch": "MoE+Hybrid", "archDetail": "Reasoning(3B/35B)", "bfcl": null, "ctxIn": 262, "ctxOut": 32, "elo": 1380, "gpqa": 83.0, "group": "qwen", "hle": 15.0, "ifeval": null, "lcb": null, "license": "Apache2", "metacog": 59.3, "mmluPro": null, "mmmlu": 78, "priceIn": 0.1, "priceOut": 0.4, "released": "2026.01", "sciCode": null, "swePro": 18.0, "sweV": 68.0, "termBench": null, "tokPerSec": 194, "ttft": 4.0, "type": "open", "unionEval": null, "url": "https://huggingface.co/Qwen", "vision": "Image, Video"}
+{"name": "Qwen3.5-Flash", "provider": "Alibaba", "aime": null, "arcAgi2": null, "arch": "MoE+Hybrid", "archDetail": "Non-Reasoning", "bfcl": null, "ctxIn": 1000, "ctxOut": 32, "elo": null, "gpqa": null, "group": "qwen", "hle": null, "ifeval": null, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": null, "mmmlu": 75, "priceIn": 0.1, "priceOut": 0.4, "released": "2026.01", "sciCode": null, "swePro": null, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": 3.0, "type": "closed", "unionEval": null, "url": "https://huggingface.co/Qwen", "vision": "Image, Video"}
+{"name": "Qwen3.5-9B", "provider": "Alibaba", "aime": null, "arcAgi2": null, "arch": "Dense", "archDetail": "Reasoning(9B)", "bfcl": 66.1, "ctxIn": 262, "ctxOut": 32, "elo": 1300, "gpqa": 81.7, "group": "qwen", "hle": 10.0, "ifeval": 91.5, "lcb": 65.6, "license": "Apache2", "metacog": null, "mmluPro": 82.5, "mmmlu": 81.2, "priceIn": null, "priceOut": null, "released": "2026.01", "sciCode": null, "swePro": null, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": 3.5, "type": "open", "unionEval": null, "url": "https://huggingface.co/Qwen", "vision": "Image, Video"}
+{"name": "Qwen3.5-4B", "provider": "Alibaba", "aime": null, "arcAgi2": null, "arch": "Dense", "archDetail": "Reasoning(4B)", "bfcl": 50.3, "ctxIn": 262, "ctxOut": 32, "elo": null, "gpqa": 76.2, "group": "qwen", "hle": null, "ifeval": 89.8, "lcb": 55.8, "license": "Apache2", "metacog": null, "mmluPro": 79.1, "mmmlu": 76.1, "priceIn": null, "priceOut": null, "released": "2026.01", "sciCode": null, "swePro": null, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": 2.0, "type": "open", "unionEval": null, "url": "https://huggingface.co/Qwen", "vision": "Image, Video"}
+{"name": "DeepSeek V3.2", "provider": "DeepSeek", "aime": 89.3, "arcAgi2": null, "arch": "MoE", "archDetail": "Non-Reasoning(37B/671B)", "bfcl": null, "ctxIn": 128, "ctxOut": 8, "elo": 1430, "gpqa": 79.9, "group": "value", "hle": 22.0, "ifeval": 87.5, "lcb": null, "license": "MIT", "metacog": 60.04, "mmluPro": null, "mmmlu": 78, "priceIn": 0.14, "priceOut": 0.28, "released": "2025.12", "sciCode": null, "swePro": 15.6, "sweV": 66.0, "termBench": 39.6, "tokPerSec": null, "ttft": 7.0, "type": "open", "unionEval": null, "url": "https://huggingface.co/deepseek-ai/DeepSeek-V3", "vision": "Text"}
+{"name": "DeepSeek R1", "provider": "DeepSeek", "aime": 87.5, "arcAgi2": null, "arch": "MoE", "archDetail": "Reasoning(37B/671B)", "bfcl": null, "ctxIn": 128, "ctxOut": 8, "elo": 1440, "gpqa": 82.0, "group": "value", "hle": 14.0, "ifeval": 83.3, "lcb": null, "license": "MIT", "metacog": 56.7, "mmluPro": 85.0, "mmmlu": 74, "priceIn": 0.55, "priceOut": 2.19, "released": "2025.01", "sciCode": null, "swePro": 18.0, "sweV": 49.2, "termBench": null, "tokPerSec": null, "ttft": 8.0, "type": "open", "unionEval": null, "url": "https://huggingface.co/deepseek-ai/DeepSeek-R1", "vision": "Text"}
+{"name": "Kimi K2.5", "provider": "Moonshot", "aime": 96.1, "arcAgi2": 12.1, "arch": "MoE", "archDetail": "Reasoning(1T)", "bfcl": null, "ctxIn": 200, "ctxOut": 32, "elo": 1447, "gpqa": 87.6, "group": "flagship", "hle": 44.9, "ifeval": 94.0, "lcb": 85.0, "license": "MIT", "metacog": 68.71, "mmluPro": 92.0, "mmmlu": 81, "priceIn": 0.55, "priceOut": 2.5, "released": "2025.11", "sciCode": null, "swePro": 27.7, "sweV": 76.8, "termBench": 43.2, "tokPerSec": null, "ttft": 5.0, "type": "open", "unionEval": null, "url": "https://huggingface.co/moonshotai", "vision": "Image, Video"}
+{"name": "GLM-5", "provider": "Zhipu AI", "aime": 92.7, "arcAgi2": null, "arch": "MoE", "archDetail": "Reasoning(40B/745B)", "bfcl": null, "ctxIn": 200, "ctxOut": 16, "elo": 1451, "gpqa": 86.0, "group": "value", "hle": 30.5, "ifeval": null, "lcb": 52.0, "license": "MIT", "metacog": 62.5, "mmluPro": 87.0, "mmmlu": 76, "priceIn": 0.35, "priceOut": 0.39, "released": "2026.02", "sciCode": null, "swePro": 77.8, "sweV": null, "termBench": 52.4, "tokPerSec": null, "ttft": null, "type": "open", "unionEval": null, "url": "https://huggingface.co/zai-org/GLM-5", "vision": "Text"}
+{"name": "Llama 4 Scout", "provider": "Meta", "aime": 85.0, "arcAgi2": null, "arch": "MoE", "archDetail": "Non-Reasoning(17B/400B)", "bfcl": null, "ctxIn": 10000, "ctxOut": 16, "elo": 1340, "gpqa": 73.0, "group": "flagship", "hle": 12.0, "ifeval": 85.0, "lcb": null, "license": "Meta", "metacog": 45.1, "mmluPro": null, "mmmlu": 70, "priceIn": 0.11, "priceOut": 0.34, "released": "2025.04", "sciCode": null, "swePro": 5.2, "sweV": 55.0, "termBench": null, "tokPerSec": null, "ttft": 2.0, "type": "open", "unionEval": null, "url": "https://huggingface.co/meta-llama", "vision": "Image, Video"}
+{"name": "Mistral Large 3", "provider": "Mistral", "aime": 82.0, "arcAgi2": null, "arch": "MoE", "archDetail": "Non-Reasoning(675B)", "bfcl": null, "ctxIn": 256, "ctxOut": 16, "elo": 1320, "gpqa": 78.0, "group": "flagship", "hle": 11.0, "ifeval": 86.0, "lcb": null, "license": "Apache2", "metacog": null, "mmluPro": null, "mmmlu": 72, "priceIn": 2.0, "priceOut": 6.0, "released": "2025.11", "sciCode": null, "swePro": 12.0, "sweV": 60.0, "termBench": null, "tokPerSec": null, "ttft": 0.3, "type": "open", "unionEval": null, "url": "https://huggingface.co/mistralai", "vision": "Image"}
+{"name": "Gemini 3 Flash", "provider": "Google", "aime": 95.0, "arcAgi2": 34.0, "arch": "Dense", "archDetail": "Non-Reasoning+Thinking", "bfcl": null, "ctxIn": 1000, "ctxOut": 64, "elo": 1490, "gpqa": 90.4, "group": "flagship", "hle": 33.7, "ifeval": 88.2, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": 88.6, "mmmlu": 83, "priceIn": 0.5, "priceOut": 3.0, "released": "2025.12", "sciCode": null, "swePro": 71.2, "sweV": 72.5, "termBench": 64.3, "tokPerSec": 218, "ttft": 1.2, "type": "closed", "unionEval": null, "url": "https://deepmind.google/technologies/gemini/flash/", "vision": "Image, Video, Audio"}
+{"name": "Llama 4 Maverick", "provider": "Meta", "aime": 82.0, "arcAgi2": null, "arch": "MoE", "archDetail": "Non-Reasoning(17B/400B)", "bfcl": null, "ctxIn": 1000, "ctxOut": 16, "elo": 1390, "gpqa": 69.8, "group": "flagship", "hle": 18.0, "ifeval": 83.0, "lcb": null, "license": "Llama4", "metacog": 44.9, "mmluPro": 80.5, "mmmlu": 74, "priceIn": 0.22, "priceOut": 0.88, "released": "2025.10", "sciCode": null, "swePro": 12.3, "sweV": 73.0, "termBench": null, "tokPerSec": null, "ttft": 4.5, "type": "open", "unionEval": null, "url": "https://huggingface.co/meta-llama", "vision": "Image, Video"}
+{"name": "Claude Haiku 4.5", "provider": "Anthropic", "aime": null, "arcAgi2": null, "arch": "Dense", "archDetail": "Non-Reasoning", "bfcl": null, "ctxIn": 200, "ctxOut": 8, "elo": 1350, "gpqa": 75.0, "group": "flagship", "hle": null, "ifeval": 86.5, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": 72.0, "mmmlu": 71, "priceIn": 1.0, "priceOut": 5.0, "released": "2025.09", "sciCode": null, "swePro": 14.0, "sweV": 68.0, "termBench": 35.5, "tokPerSec": null, "ttft": 0.4, "type": "closed", "unionEval": null, "url": "https://anthropic.com/claude", "vision": "Image"}
+{"name": "Grok 4.1 Fast", "provider": "xAI", "aime": 88.0, "arcAgi2": null, "arch": "Dense", "archDetail": "Reasoning", "bfcl": null, "ctxIn": 2000, "ctxOut": 16, "elo": 1380, "gpqa": 85.3, "group": "flagship", "hle": null, "ifeval": null, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": null, "mmmlu": 72, "priceIn": 0.2, "priceOut": 0.8, "released": "2025.11", "sciCode": null, "swePro": null, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": 1.5, "type": "closed", "unionEval": null, "url": "https://x.ai/grok", "vision": "Image, Video"}
+{"name": "Phi-4", "provider": "Microsoft", "aime": null, "arcAgi2": null, "arch": "Dense", "archDetail": "Non-Reasoning(14B)", "bfcl": null, "ctxIn": 16, "ctxOut": 4, "elo": 1310, "gpqa": 73.0, "group": "value", "hle": null, "ifeval": null, "lcb": null, "license": "MIT", "metacog": null, "mmluPro": null, "mmmlu": 68, "priceIn": null, "priceOut": null, "released": "2024.12", "sciCode": null, "swePro": null, "sweV": 72.0, "termBench": null, "tokPerSec": null, "ttft": 2.5, "type": "open", "unionEval": null, "url": "https://huggingface.co/microsoft/phi-4", "vision": "Image"}
+{"name": "K-EXAONE", "provider": "LG AI Research", "aime": 85.3, "arcAgi2": null, "arch": "MoE", "archDetail": "Reasoning(236B, K-Foundation)", "bfcl": null, "ctxIn": 260, "ctxOut": 16, "elo": null, "gpqa": 75.4, "group": "korean", "hle": null, "ifeval": null, "lcb": null, "license": "Research", "metacog": null, "mmluPro": 81.8, "mmmlu": 72, "priceIn": 0, "priceOut": 0, "released": "2025.12", "sciCode": null, "swePro": null, "sweV": 49.4, "termBench": null, "tokPerSec": null, "ttft": null, "type": "open", "unionEval": null, "url": "https://huggingface.co/LGAI-EXAONE", "vision": "Image, Text"}
+{"name": "A.X K1", "provider": "SK Telecom", "aime": null, "arcAgi2": null, "arch": "MoE", "archDetail": "Reasoning(500B, K-Foundation)", "bfcl": null, "ctxIn": 64, "ctxOut": 16, "elo": null, "gpqa": null, "group": "korean", "hle": null, "ifeval": null, "lcb": null, "license": "Apache2", "metacog": null, "mmluPro": null, "mmmlu": 62, "priceIn": 0, "priceOut": 0, "released": "2025.12", "sciCode": null, "swePro": null, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": null, "type": "closed", "unionEval": null, "url": "https://www.sktelecom.com", "vision": "Text"}
+{"name": "Solar Open 100B", "provider": "Upstage", "aime": 84.3, "arcAgi2": null, "arch": "MoE", "archDetail": "Reasoning(100B, K-Foundation)", "bfcl": null, "ctxIn": 100, "ctxOut": 16, "elo": null, "gpqa": 68.1, "group": "korean", "hle": null, "ifeval": null, "lcb": null, "license": "Apache2", "metacog": null, "mmluPro": 80.4, "mmmlu": 68, "priceIn": 0, "priceOut": 0, "released": "2025.12", "sciCode": null, "swePro": 74.2, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": null, "type": "open", "unionEval": null, "url": "https://huggingface.co/upstage/Solar-Open-100B", "vision": "Text"}
+{"name": "Motif AI", "provider": "Motif Technologies", "aime": null, "arcAgi2": null, "arch": "Dense", "archDetail": "Non-Reasoning(K-Foundation)", "bfcl": null, "ctxIn": 128, "ctxOut": 16, "elo": null, "gpqa": null, "group": "korean", "hle": null, "ifeval": null, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": null, "mmmlu": 55, "priceIn": 0, "priceOut": 0, "released": "2026.02", "sciCode": null, "swePro": null, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": null, "type": "closed", "unionEval": null, "url": "https://motif.ai", "vision": "Text"}
+{"name": "Mi:dm 2.0 Base", "provider": "KT", "aime": null, "arcAgi2": null, "arch": "Dense", "archDetail": "Non-Reasoning(11.5B, Llama-DuS)", "bfcl": null, "ctxIn": 32, "ctxOut": 4, "elo": null, "gpqa": null, "group": "korean", "hle": null, "ifeval": null, "lcb": null, "license": "MIT", "metacog": null, "mmluPro": null, "mmmlu": null, "priceIn": 0, "priceOut": 0, "released": "2025.07", "sciCode": null, "swePro": null, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": null, "type": "open", "unionEval": null, "url": "https://huggingface.co/K-intelligence/Midm-2.0-Base-Instruct", "vision": "Text"}
+{"name": "Nanbeige4.1-3B", "provider": "Nanbeige", "aime": null, "arcAgi2": null, "arch": "Dense", "archDetail": "Reasoning(3B)", "bfcl": 56.5, "ctxIn": 131, "ctxOut": 131, "elo": null, "gpqa": 83.8, "group": "value", "hle": 12.6, "ifeval": null, "lcb": 76.9, "license": "Apache2", "metacog": null, "mmluPro": null, "mmmlu": null, "priceIn": 0, "priceOut": 0, "released": "2026.03", "sciCode": null, "swePro": null, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": null, "type": "open", "unionEval": null, "url": "https://huggingface.co/Nanbeige/Nanbeige4.1-3B", "vision": "Text"}
+{"name": "MiniMax-M2.5", "provider": "MiniMax", "aime": 86.3, "arcAgi2": null, "arch": "MoE", "archDetail": "Reasoning(10B/230B)", "bfcl": null, "ctxIn": 1000, "ctxOut": 64, "elo": null, "gpqa": 84.8, "group": "flagship", "hle": 19.1, "ifeval": 87.5, "lcb": 82.6, "license": "MIT", "metacog": 60.54, "mmluPro": 82.0, "mmmlu": null, "priceIn": 0.3, "priceOut": 1.2, "released": "2026.02", "sciCode": null, "swePro": null, "sweV": 80.2, "termBench": 42.2, "tokPerSec": 47, "ttft": 3.3, "type": "open", "unionEval": null, "url": "https://huggingface.co/MiniMaxAI/MiniMax-M2.5", "vision": "Text"}
+{"name": "Step-3.5-Flash", "provider": "StepFun", "aime": 97.3, "arcAgi2": null, "arch": "MoE", "archDetail": "Reasoning(11B/196B)", "bfcl": null, "ctxIn": 256, "ctxOut": 32, "elo": null, "gpqa": null, "group": "flagship", "hle": null, "ifeval": null, "lcb": 86.4, "license": "Apache2", "metacog": null, "mmluPro": null, "mmmlu": null, "priceIn": null, "priceOut": null, "released": "2026.02", "sciCode": null, "swePro": null, "sweV": 74.4, "termBench": 51.0, "tokPerSec": 300, "ttft": null, "type": "open", "unionEval": null, "url": "https://huggingface.co/stepfun-ai/Step-3.5-Flash", "vision": "Text"}
+{"name": "GPT-5.4", "provider": "OpenAI", "aime": null, "arcAgi2": 73.3, "arch": "Dense", "archDetail": "Reasoning(Computer Use)", "bfcl": null, "ctxIn": 1000, "ctxOut": 64, "elo": null, "gpqa": 92.8, "group": "flagship", "hle": 52.1, "ifeval": null, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": null, "mmmlu": null, "priceIn": 2.5, "priceOut": 15.0, "released": "2026.03", "sciCode": null, "swePro": 57.7, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": null, "type": "closed", "unionEval": null, "url": "https://openai.com/gpt-5", "vision": "Image"}
+{"name": "GPT-5.1", "provider": "OpenAI", "aime": 94.0, "arcAgi2": 17.0, "arch": "Dense", "archDetail": "Reasoning(Adaptive)", "bfcl": null, "ctxIn": 400, "ctxOut": 64, "elo": 1480, "gpqa": 88.1, "group": "flagship", "hle": 26.0, "ifeval": null, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": null, "mmmlu": null, "priceIn": 1.25, "priceOut": 10.0, "released": "2025.11", "sciCode": null, "swePro": null, "sweV": 74.9, "termBench": null, "tokPerSec": null, "ttft": 1.0, "type": "closed", "unionEval": null, "url": "https://openai.com/gpt-5", "vision": "Image"}
+{"name": "Gemini 3 Pro", "provider": "Google", "aime": 95.0, "arcAgi2": 31.1, "arch": "Dense", "archDetail": "Reasoning(DeepThink)", "bfcl": null, "ctxIn": 1000, "ctxOut": 64, "elo": 1490, "gpqa": 91.9, "group": "flagship", "hle": 37.5, "ifeval": null, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": null, "mmmlu": 85, "priceIn": 2.0, "priceOut": 12.0, "released": "2025.11", "sciCode": null, "swePro": null, "sweV": 76.2, "termBench": null, "tokPerSec": 134, "ttft": 3.0, "type": "closed", "unionEval": null, "url": "https://deepmind.google/technologies/gemini/", "vision": "Image, Video, Audio"}
+{"name": "Claude Sonnet 4.5", "provider": "Anthropic", "aime": 100, "arcAgi2": null, "arch": "Dense", "archDetail": "Reasoning(Hybrid)", "bfcl": null, "ctxIn": 200, "ctxOut": 64, "elo": 1440, "gpqa": 83.4, "group": "flagship", "hle": 30.8, "ifeval": 90.0, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": null, "mmmlu": null, "priceIn": 3.0, "priceOut": 15.0, "released": "2025.09", "sciCode": null, "swePro": null, "sweV": 77.2, "termBench": 51.0, "tokPerSec": null, "ttft": 1.5, "type": "closed", "unionEval": null, "url": "https://anthropic.com/claude", "vision": "Image"}
+{"name": "GPT-5-Nano", "provider": "OpenAI", "aime": null, "arcAgi2": null, "arch": "Dense", "archDetail": "Non-Reasoning(Nano)", "bfcl": null, "ctxIn": 400, "ctxOut": 16, "elo": null, "gpqa": null, "group": "value", "hle": null, "ifeval": null, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": null, "mmmlu": null, "priceIn": 0.05, "priceOut": 0.4, "released": "2025.08", "sciCode": null, "swePro": null, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": 0.2, "type": "closed", "unionEval": null, "url": "https://openai.com/gpt-5", "vision": "Image"}
+{"name": "Gemini 2.5 FL-Lite", "provider": "Google", "aime": null, "arcAgi2": null, "arch": "Dense", "archDetail": "Non-Reasoning+Thinking", "bfcl": null, "ctxIn": 1000, "ctxOut": 64, "elo": null, "gpqa": null, "group": "value", "hle": null, "ifeval": null, "lcb": null, "license": "Prop", "metacog": null, "mmluPro": null, "mmmlu": null, "priceIn": 0.1, "priceOut": 0.4, "released": "2025.06", "sciCode": null, "swePro": null, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": 0.5, "type": "closed", "unionEval": null, "url": "https://deepmind.google/technologies/gemini/flash/", "vision": "Image, Video, Audio"}
+{"name": "Qwen3-Next-80B", "provider": "Alibaba", "aime": null, "arcAgi2": null, "arch": "MoE+Hybrid", "archDetail": "Reasoning(3B/80B)", "bfcl": 49.7, "ctxIn": 262, "ctxOut": 32, "elo": null, "gpqa": 77.2, "group": "qwen", "hle": null, "ifeval": 88.9, "lcb": 68.7, "license": "NCML", "metacog": null, "mmluPro": 82.7, "mmmlu": 81.3, "priceIn": 0.15, "priceOut": 1.2, "released": "2025.09", "sciCode": null, "swePro": null, "sweV": null, "termBench": null, "tokPerSec": null, "ttft": null, "type": "open", "unionEval": null, "url": "https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Thinking", "vision": "Text"}

data/music.jsonl

@@ -0,0 +1,8 @@
+{"name": "Suno v4.5", "provider": "Suno", "released": "2025.11", "quality": "S", "vocals": "S", "instrumental": "S", "lyrics": "S", "duration": "4min", "price": "$$", "license": "Prop"}
+{"name": "Udio v2", "provider": "Udio", "released": "2025.10", "quality": "S", "vocals": "A", "instrumental": "S", "lyrics": "A", "duration": "4min", "price": "$$", "license": "Prop"}
+{"name": "Gemini Music", "provider": "Google", "released": "2026.01", "quality": "A", "vocals": "A", "instrumental": "A", "lyrics": "A", "duration": "60s", "price": "$", "license": "Prop"}
+{"name": "MusicGen Large", "provider": "Meta", "released": "2024.06", "quality": "B", "vocals": "A", "instrumental": "B", "lyrics": "A", "duration": "30s", "price": "Free", "license": "Open"}
+{"name": "Stable Audio 2.0", "provider": "Stability AI", "released": "2024.04", "quality": "B", "vocals": "A", "instrumental": "B", "lyrics": "B", "duration": "3min", "price": "$", "license": "Open"}
+{"name": "JASCO", "provider": "Meta", "released": "2024.12", "quality": "C", "vocals": "A", "instrumental": "C", "lyrics": "B", "duration": "30s", "price": "Free", "license": "Open"}
+{"name": "Riffusion v2", "provider": "Riffusion", "released": "2025.06", "quality": "B", "vocals": "B", "instrumental": "B", "lyrics": "A", "duration": "30s", "price": "$", "license": "Prop"}
+{"name": "Loudme", "provider": "Loudme", "released": "2025.09", "quality": "A", "vocals": "A", "instrumental": "A", "lyrics": "B", "duration": "5min", "price": "$", "license": "Prop"}

data/video.jsonl

@@ -0,0 +1,10 @@
+{"name": "Sora 2", "provider": "OpenAI", "released": "2025.09", "quality": "S", "motion": "S", "consistency": "A", "textRender": "S", "duration": "20s", "resolution": "1080p", "price": "$$$", "license": "Prop"}
+{"name": "Veo 3.1", "provider": "Google", "released": "2026.01", "quality": "S", "motion": "A", "consistency": "S", "textRender": "S", "duration": "8s", "resolution": "4K", "price": "$$", "license": "Prop"}
+{"name": "Runway Gen-4.5", "provider": "Runway", "released": "2026.02", "quality": "S", "motion": "S", "consistency": "A", "textRender": "A", "duration": "16s", "resolution": "1080p", "price": "$$", "license": "Prop"}
+{"name": "Kling 3.0", "provider": "Kuaishou", "released": "2026.02", "quality": "A", "motion": "A", "consistency": "A", "textRender": "A", "duration": "5min", "resolution": "1080p", "price": "$", "license": "Prop"}
+{"name": "Seedance 2.0", "provider": "ByteDance", "released": "2026.01", "quality": "S", "motion": "A", "consistency": "S", "textRender": "A", "duration": "10s", "resolution": "2K", "price": "$$", "license": "Prop"}
+{"name": "Wan 2.6", "provider": "Alibaba", "released": "2026.01", "quality": "A", "motion": "A", "consistency": "B", "textRender": "A", "duration": "10s", "resolution": "1080p", "price": "Free", "license": "Open"}
+{"name": "Pika 2.5", "provider": "Pika", "released": "2025.12", "quality": "B", "motion": "B", "consistency": "B", "textRender": "A", "duration": "10s", "resolution": "1080p", "price": "$", "license": "Prop"}
+{"name": "Luma Ray3", "provider": "Luma", "released": "2026.01", "quality": "A", "motion": "A", "consistency": "B", "textRender": "B", "duration": "10s", "resolution": "4K", "price": "$$", "license": "Prop"}
+{"name": "LTX-2", "provider": "Lightricks", "released": "2026.01", "quality": "A", "motion": "B", "consistency": "A", "textRender": "B", "duration": "10s", "resolution": "4K", "price": "Free", "license": "Open"}
+{"name": "HaiLuo AI", "provider": "MiniMax", "released": "2025.12", "quality": "B", "motion": "A", "consistency": "B", "textRender": "B", "duration": "6s", "resolution": "1080p", "price": "$", "license": "Prop"}

data/vlm_flagship.jsonl

@@ -0,0 +1,11 @@
+{"name": "Gemini 3 Flash", "provider": "Google", "mmmu": 87.6, "mmmuPro": 80.0, "mathVista": null, "ai2d": null, "ocrBench": null, "mmStar": null, "hallusionBench": null, "mmBenchEN": null, "realWorldQA": null, "videoMME": null}
+{"name": "Gemini 3 Pro", "provider": "Google", "mmmu": 87.5, "mmmuPro": 80.0, "mathVista": null, "ai2d": null, "ocrBench": null, "mmStar": null, "hallusionBench": null, "mmBenchEN": null, "realWorldQA": null, "videoMME": null}
+{"name": "Gemini 3.1 Pro", "provider": "Google", "mmmu": null, "mmmuPro": 82.0, "mathVista": null, "ai2d": null, "ocrBench": null, "mmStar": null, "hallusionBench": null, "mmBenchEN": null, "realWorldQA": null, "videoMME": null}
+{"name": "GPT-5.2", "provider": "OpenAI", "mmmu": 86.7, "mmmuPro": null, "mathVista": null, "ai2d": null, "ocrBench": null, "mmStar": null, "hallusionBench": null, "mmBenchEN": null, "realWorldQA": null, "videoMME": null}
+{"name": "GPT-5 (original)", "provider": "OpenAI", "mmmu": 84.2, "mmmuPro": null, "mathVista": null, "ai2d": null, "ocrBench": null, "mmStar": null, "hallusionBench": null, "mmBenchEN": null, "realWorldQA": null, "videoMME": null}
+{"name": "Claude Opus 4.6", "provider": "Anthropic", "mmmu": null, "mmmuPro": 85.1, "mathVista": null, "ai2d": null, "ocrBench": null, "mmStar": null, "hallusionBench": null, "mmBenchEN": null, "realWorldQA": null, "videoMME": null}
+{"name": "Grok 4 Heavy", "provider": "xAI", "mmmu": 76.5, "mmmuPro": null, "mathVista": null, "ai2d": null, "ocrBench": null, "mmStar": null, "hallusionBench": null, "mmBenchEN": null, "realWorldQA": null, "videoMME": null}
+{"name": "InternVL3.5-241B", "provider": "OpenGVLab", "mmmu": 77.7, "mmmuPro": null, "mathVista": null, "ai2d": null, "ocrBench": null, "mmStar": null, "hallusionBench": null, "mmBenchEN": null, "realWorldQA": null, "videoMME": null}
+{"name": "InternVL3-78B", "provider": "OpenGVLab", "mmmu": 72.2, "mmmuPro": null, "mathVista": 79.6, "ai2d": 89.7, "ocrBench": 90.6, "mmStar": null, "hallusionBench": 59.1, "mmBenchEN": 89.0, "realWorldQA": 78.0, "videoMME": null}
+{"name": "Qwen2.5-VL-72B", "provider": "Alibaba", "mmmu": 70.2, "mmmuPro": null, "mathVista": 74.8, "ai2d": null, "ocrBench": null, "mmStar": 70.8, "hallusionBench": null, "mmBenchEN": null, "realWorldQA": null, "videoMME": null}
+{"name": "Kimi-VL-A3B-Thinking", "provider": "Moonshot", "mmmu": 64.0, "mmmuPro": 46.3, "mathVista": 80.1, "ai2d": null, "ocrBench": null, "mmStar": 70.4, "hallusionBench": null, "mmBenchEN": 84.4, "realWorldQA": 70.0, "videoMME": null}

schema.jsonld

@@ -0,0 +1,139 @@
+{
+  "@context": "https://schema.org",
+  "@type": "Dataset",
+  "name": "ALL Bench Leaderboard 2026",
+  "alternateName": [
+    "ALL Bench",
+    "ALLBench",
+    "AI Benchmark Leaderboard 2026"
+  ],
+  "description": "The only AI benchmark leaderboard covering LLM, VLM, Agent, Image, Video, and Music generation in a single unified view. 91 models cross-verified across 6 modalities with confidence badges. Features composite 5-axis scoring, interactive comparison tools, and downloadable intelligence reports.",
+  "url": "https://huggingface.co/spaces/FINAL-Bench/all-bench-leaderboard",
+  "sameAs": [
+    "https://huggingface.co/datasets/FINAL-Bench/ALL-Bench-Leaderboard",
+    "https://github.com/final-bench/ALL-Bench-Leaderboard"
+  ],
+  "license": "https://opensource.org/licenses/MIT",
+  "version": "2.2.1",
+  "datePublished": "2026-03-01",
+  "dateModified": "2026-03-10",
+  "creator": {
+    "@type": "Organization",
+    "name": "ALL Bench Team",
+    "url": "https://huggingface.co/FINAL-Bench"
+  },
+  "keywords": [
+    "AI benchmark",
+    "LLM leaderboard",
+    "GPT-5",
+    "Claude",
+    "Gemini",
+    "VLM benchmark",
+    "AI agent",
+    "image generation",
+    "video generation",
+    "music generation",
+    "MMLU-Pro",
+    "GPQA",
+    "ARC-AGI-2",
+    "FINAL Bench",
+    "metacognition",
+    "multimodal AI",
+    "AI evaluation",
+    "benchmark comparison",
+    "AI model ranking",
+    "open source AI"
+  ],
+  "about": [
+    {
+      "@type": "Thing",
+      "name": "Large Language Model"
+    },
+    {
+      "@type": "Thing",
+      "name": "Vision Language Model"
+    },
+    {
+      "@type": "Thing",
+      "name": "AI Benchmark"
+    },
+    {
+      "@type": "Thing",
+      "name": "Generative AI"
+    },
+    {
+      "@type": "Thing",
+      "name": "Metacognition"
+    }
+  ],
+  "measurementTechnique": "Cross-verified benchmark aggregation with 3-tier confidence system",
+  "variableMeasured": [
+    {
+      "@type": "PropertyValue",
+      "name": "MMLU-Pro",
+      "description": "57K expert-level multi-discipline questions"
+    },
+    {
+      "@type": "PropertyValue",
+      "name": "GPQA Diamond",
+      "description": "PhD-level expert questions in science"
+    },
+    {
+      "@type": "PropertyValue",
+      "name": "AIME 2025",
+      "description": "American Invitational Mathematics Examination"
+    },
+    {
+      "@type": "PropertyValue",
+      "name": "HLE",
+      "description": "Humanity's Last Exam — 2500 expert-sourced questions"
+    },
+    {
+      "@type": "PropertyValue",
+      "name": "ARC-AGI-2",
+      "description": "Abstract reasoning and novel pattern recognition"
+    },
+    {
+      "@type": "PropertyValue",
+      "name": "FINAL Bench Metacognitive",
+      "description": "AI self-correction ability measurement"
+    },
+    {
+      "@type": "PropertyValue",
+      "name": "SWE-Pro",
+      "description": "Software engineering benchmark by Scale AI"
+    },
+    {
+      "@type": "PropertyValue",
+      "name": "IFEval",
+      "description": "Instruction following evaluation"
+    },
+    {
+      "@type": "PropertyValue",
+      "name": "LiveCodeBench",
+      "description": "Continuously updated coding benchmark"
+    },
+    {
+      "@type": "PropertyValue",
+      "name": "Union Eval S3",
+      "description": "ALL Bench proprietary integrated benchmark, 100% JSON auto-graded"
+    }
+  ],
+  "distribution": [
+    {
+      "@type": "DataDownload",
+      "encodingFormat": "application/json",
+      "contentUrl": "https://huggingface.co/datasets/FINAL-Bench/ALL-Bench-Leaderboard/resolve/main/all_bench_leaderboard_v2.1.json",
+      "name": "Unified JSON Dataset (75KB)"
+    }
+  ],
+  "isPartOf": {
+    "@type": "DataCatalog",
+    "name": "Hugging Face Datasets",
+    "url": "https://huggingface.co/datasets"
+  },
+  "funder": {
+    "@type": "Organization",
+    "name": "FINAL Bench"
+  }
+}