OpenCompass LLM - Math (CompassBench 2407): leaderboard
OpenCompass CompassBench 2407 edition (file dated 2024-08-05): math category score. CompassBench is OpenCompass's periodic evaluation of chat models on a private question set; each edition is a separate board. Score 0-100, higher is better.
Metric: Score (%). Source: rank.opencompass.org.cn. Saturation forecast: Estimated already saturated. 36 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Mistral Large 2 (Jul) | 72.8 | #490 |
| 2 | Claude 3.5 Sonnet (20240620) | 71.1 | #308 |
| 3 | GPT-4o (2024-05-13) | 71.1 | #337 |
| 4 | DeepSeek V2 Chat (0628) | 68.2 | #409 |
| 5 | Mistral Large | 66.4 | #713 |
| 6 | GPT-4o Mini (2024-07-18) | 58.2 | #568 |
| 7 | Llama 3.1 70B Instruct | 58 | #567 |
| 8 | Qwen 2 72B Instruct | 57.7 | #579 |
| 9 | Yi Large | 54.8 | #566 |
| 10 | GLM-4 (0520) | 53.2 | #612 |
| 11 | Gemma 2 27B (IT) | 50.1 | #652 |
| 12 | Mixtral 8x22B Instruct | 47.2 | #738 |
| 13 | Moonshot v1 8K | 46.6 | #622 |
| 14 | InternLM2.5-Chat-7B | 40.8 | #1009 |
| 15 | Gemma 2 9B (IT) | 40.7 | #795 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=opencompass-llm-math-compassbench-2407 · How It Works · Data refreshed daily, snapshot 2026-11-05.