Benchmarks · composite
Arena Text · Industry: mathematics
Crowdsourced pairwise human preference on open-ended text prompts (style control on); Bradley-Terry rating. Only the Industry: mathematics category.
207 models · 218 results · updated 2026-11-09 · arena.ai ↗
Leaderboard
Gemini 4 Argon high
All results
218 rows| # | Model | Effort | Score |
|---|---|---|---|
| 1 | Gemini 4 Argon | high | 1539 ±18 |
| 2 | Fable 5.1 | max | 1534 ±18 |
| 3 | Opus 5 | high | 1531 ±10 |
| 4 | Opus 5.5 | high | 1527 ±21 |
| 5 | Opus 4.6 | high | 1524 ±10 |
| 6 | Gemini 4 Carbon | high | 1523 ±40 |
| 7 | Opus 5 | max | 1522 ±14 |
| 8 | Fable 5 | high | 1520 ±12 |
| 9 | GPT-6.1 Sol | max | 1518 ±22 |
| 10 | Opus 4.7 | high | 1514 ±11 |
| 11 | Gemini 3.8 Flash | high | 1512 ±11 |
| 12 | Kimi K3 | max | 1511 ±13 |
| 13 | Opus 4.6 | max | 1511 ±9 |
| 14 | GPT-5.6 Sol | xhigh | 1510 ±13 |
| 15 | Sonnet 5.5 | xhigh | 1505 ±23 |
| 16 | MiMo V2.6 Pro | max | 1505 ±39 |
| 17 | GPT-5.5 | high | 1505 ±10 |
| 18 | Opus 4.8 | high | 1504 ±11 |
| 19 | GPT-6 Astra | max | 1504 ±19 |
| 20 | GLM 5.3 | max | 1503 ±15 |
| 21 | Muse Wonder | max | 1503 ±23 |
| 22 | Muse Spark 1.3 | max | 1500 ±16 |
| 23 | GPT-5.5 | xhigh | 1500 ±10 |
| 24 | GPT-5.4 | high | 1499 ±11 |
| 25 | Muse Spark 1.1 | max | 1497 ±12 |
| 26 | Opus 4.7 | max | 1497 ±10 |
| 27 | Gemini 3.7 Flash | high | 1496 ±14 |
| 28 | Qwen 3.8 Max | max | 1496 ±14 |
| 29 | GLM 5.2 | max | 1493 ±11 |
| 30 | MiMo V2.5 Pro | xhigh | 1492 ±10 |