OpenCompass Code - Execution - Chinese (CompassBench 2501): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 40 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O3 Mini (2025-01-31) | 31.1 |
| 2 | O1 (2024-12-17) | 30.4 |
| 3 | DeepSeek R1 | 28.4 |
| 4 | Qwen 2.5 Max | 25.7 |
| 5 | Gemini 2.0 Pro (Preview 02-05) | 25 |
| 6 | GLM-4 Plus | 23.6 |
| 7 | Qwen 2.5 72B Instruct | 23 |
| 8 | O1 Mini (2024-09-12) | 23 |
| 9 | Claude 3.5 Sonnet (20241022) | 22.3 |
| 10 | GPT-4o Mini (2024-07-18) | 22.3 |
| 11 | GPT-4o (2024-11-20) | 21.6 |
| 12 | Llama 3.1 405B Instruct FP8 | 20.9 |
| 13 | DeepSeek V3 | 20.3 |
| 14 | Doubao-1.5-pro-32k-250115 | 20.3 |
| 15 | Gemini 2.0 Flash (001) | 19.6 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-code-execution-chinese-compassbench-2501 · How It Works · Data refreshed daily, snapshot 2026-09-19.