OpenCompass Math - Reasoning Computation - Chinese (CompassBench 2501): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 40 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O3 Mini (2025-01-31) | 60.6 |
| 2 | O1 (2024-12-17) | 56.9 |
| 3 | DeepSeek R1 | 53.1 |
| 4 | O1 Mini (2024-09-12) | 41.2 |
| 5 | QwQ 32B-Preview | 31.9 |
| 6 | GLM-Zero-Preview | 31.9 |
| 7 | Gemini 2.0 Pro (Preview 02-05) | 28.8 |
| 8 | Gemini 2.0 Flash (001) | 28.1 |
| 9 | Doubao-1.5-pro-32k-250115 | 28.1 |
| 10 | DeepSeek V3 | 19.4 |
| 11 | Qwen 2.5 Max | 16.9 |
| 12 | Claude 3.5 Sonnet (20241022) | 12.5 |
| 13 | GPT-4o Mini (2024-07-18) | 11.9 |
| 14 | InternLM3-8B-Instruct | 9.4 |
| 15 | Qwen 2.5 72B Instruct | 8.8 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-math-reasoning-computation-chinese-compassbench-2501 · How It Works · Data refreshed daily, snapshot 2026-09-19.