SEA-HELM (English) - MathArena: leaderboard
Metric: Normalized Score. Source: leaderboard.sea-lion.ai. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemma 4 31B (IT) | 37.76 |
| 2 | Gemma 4 26B A4B (IT) | 29.27 |
| 3 | Gemma 4 12B (IT) | 26.54 |
| 4 | Gemma 3 27B (IT) | 14.94 |
| 5 | Qwen 3.5 27B | 13.98 |
| 6 | Qwen 3.5 122B A10B | 13.57 |
| 7 | Qwen 3 VL 32B Instruct | 12.62 |
| 8 | Qwen 3.5 35B A3B | 12.57 |
| 9 | Qwen 3.6 35B A3B | 11.96 |
| 10 | gemma-4-E4B-it | 11.87 |
| 11 | Qwen 3.6 27B | 11.6 |
| 12 | Llama 4 Scout Instruct | 11.33 |
| 13 | Gemma 3 12B (IT) | 10.93 |
| 14 | Qwen 3 VL 8B Instruct | 10.2 |
| 15 | Mistral Small 4 | 10 |
Interactive version: theaggregate.ai/benchmark?slug=sea-helm-english-matharena · How It Works · Data refreshed daily, snapshot 2026-09-19.