Document Arena
Professional Work · 2026-06-10
Arena-style leaderboard for document analysis and long-content reasoning tasks, reported as arena/Elo-style model scores.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Opus 5 | 1516 |
| Claude Fable 5.1 | 1513 |
| Opus 4.6 | 1507 |
| Claude Fable 5 | 1496 |
| Opus 4.7 | 1495 |
| GPT-5.5 | 1486 |
| GPT-5.6 Sol | 1483 |
| Sonnet 4.6 | 1482 |
| Opus 4.8 | 1475 |
| GPT-5.6 Terra | 1472 |
| GPT-5.4 | 1471 |
| Muse Spark 1.3 | 1471 |
| GPT-6 Astra | 1468 |
| Sonnet 5 | 1466 |
| Muse Spark 1.1 | 1465 |
| Gemini 3.5 Flash | 1463 |
| Opus 4.5 | 1462 |
| GPT-5.6 Luna | 1457 |
| Gemini 3.6 Flash | 1456 |
| Grok 4.6 | 1452 |
| Grok 4.5 | 1452 |
| Kimi K2.6 | 1451 |
| Sonnet 4.5 | 1450 |
| Muse Spark | 1444 |
| Qwen3.7-Plus (source unspecified) | 1444 |
| Gemini 3.1 Pro Preview | 1444 |
| MiniMax M3 | 1435 |
| Gemini 3 Pro Preview | 1434 |
| Kimi K2.5 | 1430 |
| Gemma 4 31B IT | 1425 |
| Gemini 2.5 Pro | 1421 |
| Haiku 4.5 | 1420 |
| GLM 5V Turbo | 1416 |
| Gemini 3 Flash Preview | 1413 |
| GPT-5.2 | 1405 |
| GPT-5.1 | 1403 |