OpenCompass LLM - Average - Chinese (CompassBench 2501): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 40 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek R1 | 72.9 |
| 2 | O1 (2024-12-17) | 68.4 |
| 3 | O3 Mini (2025-01-31) | 66.1 |
| 4 | Gemini 2.0 Pro (Preview 02-05) | 63 |
| 5 | Gemini 2.0 Flash (001) | 60.7 |
| 6 | Doubao-1.5-pro-32k-250115 | 60.2 |
| 7 | O1 Mini (2024-09-12) | 59.9 |
| 8 | Qwen 2.5 Max | 57.5 |
| 9 | DeepSeek V3 | 56.9 |
| 10 | GPT-4o (2024-11-20) | 54.2 |
| 11 | QwQ 32B-Preview | 53.9 |
| 12 | GLM-4 Plus | 52.1 |
| 13 | Yi Lightning | 51.8 |
| 14 | Claude 3.5 Sonnet (20241022) | 51.6 |
| 15 | GLM-Zero-Preview | 50.8 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-llm-average-chinese-compassbench-2501 · How It Works · Data refreshed daily, snapshot 2026-09-19.