Chatbot Arena (Text - Math): leaderboard
The arena.ai text board restricted to prompts that require mathematics to be applied, such as calculation, algebraic manipulation or geometric reasoning on a well-defined problem. A language model classifier applies the tag, and anonymous pairwise votes decide each battle. The Bradley-Terry rating is refit over these battles alone, so confidence intervals are wider than on the overall board, and higher is better.
Metric: Arena Score. Source: arena.ai. 400 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 4 Argon (High) | 1528 | #8 (Gemini 4 Argon) |
| 2 | Claude Fable 5 (High) | 1524 | #16 (Claude Fable 5) |
| 3 | Gemini 3.8 Flash (High) | 1522 | #21 (Gemini 3.8 Flash) |
| 4 | Claude Fable 5.1 (Max) | 1522 | #9 (Claude Fable 5.1) |
| 5 | Claude Opus 5 (High) | 1521 | #12 (Claude Opus 5) |
| 6 | Claude Opus 4.6 (High) | 1518 | #63 (Claude Opus 4.6) |
| 7 | Claude Opus 5 (Max) | 1514 | #12 (Claude Opus 5) |
| 8 | Claude Sonnet 5.5 (xHigh) | 1512 | #11 (Claude Sonnet 5.5) |
| 9 | Claude Opus 4.6 | 1507 | #63 |
| 10 | Claude Opus 4.7 (High) | 1504 | #49 (Claude Opus 4.7) |
| 11 | Gemini 3.7 Flash (High) | 1503 | #23 (Gemini 3.7 Flash) |
| 12 | GLM-5.3 Flash | 1502 | #65 |
| 13 | Gemini 3.6 Flash (High) | 1502 | #51 (Gemini 3.6 Flash) |
| 14 | Muse Spark 1.3 (Max) | 1502 | #24 (Muse Spark 1.3) |
| 15 | GPT-5.5 | 1500 | #31 |
Interactive version: theaggregate.ai/benchmark?slug=chatbot-arena-text-math · How It Works · Data refreshed daily, snapshot 2026-10-20.