Text · Factuality · Legal & Government
Arena · 50 models
| Rank | Model | Score |
|---|---|---|
| 1 | Gemini 4 Argon (High) | 1,530.0557 |
| 2 | Claude Opus 5 (High) | 1,523.9399 |
| 3 | Claude Opus 5 (Max) | 1,519.9977 |
| 4 | GPT-5.5 (High) | 1,518.5705 |
| 5 | Claude Opus 4.6 (High) | 1,516.8445 |
| 6 | Claude Opus 4.6 | 1,512.8629 |
| 7 | Claude Fable 5.1 (Max) | 1,511.5954 |
| 8 | GPT-5.4 (High) | 1,511.1984 |
| 9 | Claude Opus 5.5 (High) | 1,509.3338 |
| 10 | Claude Fable 5 (High) | 1,509.0483 |
| 11 | Claude Opus 4.7 (High) | 1,508.7411 |
| 12 | Kimi K3 (Max) | 1,505.0443 |
| 13 | Gemini 3 Pro | 1,504.988 |
| 14 | GPT-5.6 Sol (xHigh) | 1,504.5873 |
| 15 | GPT-5.5 | 1,502.3784 |
| 16 | Gemini 3.5 Flash (High) | 1,501.0071 |
| 17 | Gemini 3.7 Flash (High) | 1,498.6666 |
| 18 | GPT-5.4 | 1,498.468 |
| 19 | Qwen3.5-Max-Preview | 1,498.2815 |
| 20 | Claude Opus 4.7 | 1,497.5302 |
| 21 | GPT-6 Astra (Max) | 1,497.1345 |
| 22 | GPT-5.6 Terra (xHigh) | 1,496.3497 |
| 23 | Gemini 3.1 Pro Preview | 1,495.2859 |
| 24 | Claude Opus 4.8 (High) | 1,495.0987 |
| 25 | Muse Spark 1.3 (Max) | 1,494.4455 |
| 26 | Claude Opus 4.8 | 1,493.0672 |
| 27 | Qwen3.8-Max | 1,491.5144 |
| 28 | Gemini 3.5 Flash (Medium) | 1,491.1327 |
| 29 | Gemini 3.8 Flash (High) | 1,490.3079 |
| 30 | GLM-5.2 (Max) | 1,489.574 |
| 31 | Gemini 2.5 Pro | 1,486.6485 |
| 32 | Grok 4.5 | 1,485.5546 |
| 33 | GPT-5.6 Luna (xHigh) | 1,485.5166 |
| 34 | Gemini 3.6 Flash (High) | 1,483.6971 |
| 35 | Claude Opus 4.5 (20251101) | 1,482.6383 |
| 36 | Muse Spark 1.1 | 1,482.0148 |
| 37 | MiMo-V2.5-Pro | 1,481.5536 |
| 38 | Claude Sonnet 4.6 | 1,481.3128 |
| 39 | GPT-5.1 (High) | 1,480.5669 |
| 40 | DeepSeek-V4-Pro | 1,478.1315 |
| 41 | GLM-5.3 (Max) | 1,478.0995 |
| 42 | GPT-5.1 | 1,477.559 |
| 43 | Claude Opus 4.5 (High, 32K, 20251101) | 1,477.3651 |
| 44 | GLM-5.1 | 1,476.8769 |
| 45 | DeepSeek-V4-Pro (High, 20260813) | 1,475.9281 |
| 46 | Claude Sonnet 5 (High) | 1,475.653 |
| 47 | Gemini 3 Flash | 1,474.8709 |
| 48 | GLM-5.3-Flash | 1,472.4224 |
| 49 | Kimi K2.6 | 1,471.9581 |
| 50 | DeepSeek-V4-Pro Preview (High) | 1,470.709 |