OpenEval - Omni-MATH: leaderboard
Human-Centered Eval OpenEval accuracy on Omni-MATH olympiad-style mathematical reasoning problems.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 1.5 32B Chat | 14.3 | #2379 |
| 2 | Qwen 1.5 14B Chat | 12.03 | #2974 |
| 3 | Qwen 1.5 7B Chat | 9.53 | #6361 |
| 4 | Qwen-14B-Chat | 7.72 | #4967 |
| 5 | Qwen 1.5 4B Chat | 6.33 | #8077 |
| 6 | Qwen 2 1.5B Instruct | 5.23 | #7855 |
| 7 | Qwen 1.5 1.8B Chat | 2.23 | #9053 |
| 8 | Qwen 1.5 0.5B Chat | 0.93 | #9551 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=openeval-omni-math · How It Works · Data refreshed daily, snapshot 2026-11-13.