Text · Style Control · Creative Writing
Arena · 50 models
| Rank | Model | Score |
|---|---|---|
| 1 | Claude Opus 5.5 (High) | 1,520.1813 |
| 2 | Gemini 4 Argon (High) | 1,517.0288 |
| 3 | Gemini 4 Carbon (High) | 1,514.9369 |
| 4 | Claude Fable 5 (High) | 1,500.6844 |
| 5 | Claude Opus 4.6 (High) | 1,499.2316 |
| 6 | Gemini 3.7 Flash (High) | 1,489.5779 |
| 7 | Claude Opus 4.7 (High) | 1,488.9421 |
| 8 | Claude Fable 5.1 (Max) | 1,488.7032 |
| 9 | Muse Wonder (Max) | 1,487.2436 |
| 10 | Gemini 3.8 Flash (High) | 1,487.0368 |
| 11 | Gemini 3 Pro | 1,484.5107 |
| 12 | Claude Opus 4.7 | 1,482.5515 |
| 13 | Gemini 3.1 Pro Preview | 1,481.3573 |
| 14 | Claude Opus 4.6 | 1,477.4676 |
| 15 | Claude Opus 5 (Max) | 1,473.761 |
| 16 | Claude Opus 5 (High) | 1,472.2683 |
| 17 | Gemini 3.6 Flash (High) | 1,471.4346 |
| 18 | GPT-5.6 Sol (xHigh) | 1,470.0178 |
| 19 | Gemini 3.5 Flash (Medium) | 1,469.7988 |
| 20 | Claude Opus 4.5 (High, 32K, 20251101) | 1,469.5559 |
| 21 | Claude Opus 4.8 (High) | 1,469.3087 |
| 22 | Qwen3.8-Max | 1,468.5663 |
| 23 | Claude Sonnet 5.5 (xHigh) | 1,468.4189 |
| 24 | Gemini 3.5 Flash (High) | 1,464.823 |
| 25 | Muse Spark 1.3 (Max) | 1,464.3428 |
| 26 | Muse Spark | 1,464.0469 |
| 27 | Grok 4.20 Beta 1 | 1,463.3837 |
| 28 | Claude Opus 4.8 | 1,463.2061 |
| 29 | Kimi K3 (Max) | 1,462.6792 |
| 30 | Claude Opus 4.5 (20251101) | 1,461.2491 |
| 31 | Gemini 3 Flash | 1,456.7314 |
| 32 | GLM-5.3 (Max) | 1,456.6088 |
| 33 | GPT-5.5 Instant | 1,454.2179 |
| 34 | Claude Sonnet 4.5 (20250929) | 1,454.0217 |
| 35 | GPT-6.1 Sol (Max) | 1,453.9205 |
| 36 | GLM-5.2 (Max) | 1,452.1461 |
| 37 | Grok 4.5 | 1,451.6715 |
| 38 | GPT-5.5 (High) | 1,451.4098 |
| 39 | Claude Sonnet 4.5 (High, 32K, 20250929) | 1,449.6705 |
| 40 | GPT-6 Astra (Max) | 1,449.5804 |
| 41 | Claude Sonnet 4.6 | 1,449.2989 |
| 42 | GPT-5.5 | 1,448.4243 |
| 43 | GLM-5 | 1,447.6416 |
| 44 | Grok 4.20 Multi-Agent Beta (0309) | 1,447.1383 |
| 45 | Gemini 3 Flash (Minimal) | 1,446.9426 |
| 46 | DeepSeek-V4-Pro (High, 20260813) | 1,446.6977 |
| 47 | DeepSeek-V4-Pro | 1,446.4591 |
| 48 | GLM-5.1 | 1,445.7774 |
| 49 | Grok 4.20 Beta (Reasoning, 0309) | 1,445.3948 |
| 50 | Claude Opus 4.1 (Thinking 16K, 20250805) | 1,445.3871 |