OpenCompass Math - Arithmetic (CompassBench 2405): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 39 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude 3 Opus | 82.7 |
| 2 | DeepSeek V2 Chat | 80.7 |
| 3 | GPT-4o (2024-05-13) | 76.3 |
| 4 | GPT-4 Turbo | 75 |
| 5 | Mixtral 8x22B Instruct (v0.1) | 74 |
| 6 | Yi-1.5-9B Chat | 73.7 |
| 7 | Command-R+ | 73 |
| 8 | Yi 1.5 34B Chat | 71 |
| 9 | GPT-3.5 Turbo | 70.7 |
| 10 | Llama 3 70B Instruct | 70 |
| 11 | Yi Large | 69.7 |
| 12 | Moonshot v1 8K | 69 |
| 13 | Qwen 1.5 110B Chat | 68.7 |
| 14 | Mistral Large | 68 |
| 15 | DeepSeek-V2-Lite-Chat | 66.7 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-math-arithmetic-compassbench-2405 · How It Works · Data refreshed daily, snapshot 2026-09-19.