OpenCompass LLM - Language (CompassBench 2501): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 40 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek R1 | 75.6 |
| 2 | Qwen 2.5 Max | 75.5 |
| 3 | Gemini 2.0 Pro (Preview 02-05) | 73.5 |
| 4 | O1 (2024-12-17) | 70.2 |
| 5 | GPT-4o (2024-11-20) | 69.5 |
| 6 | DeepSeek V3 | 68.2 |
| 7 | MiniMax-Text-01 | 68.1 |
| 8 | Gemini 2.0 Flash (001) | 67.3 |
| 9 | O3 Mini (2025-01-31) | 66.9 |
| 10 | Yi Lightning | 66.9 |
| 11 | O1 Mini (2024-09-12) | 66.5 |
| 12 | GLM-4 Plus | 62.1 |
| 13 | InternLM3-8B-Instruct | 61.5 |
| 14 | Qwen 2.5 72B Instruct | 59.6 |
| 15 | Mistral Large 2 (Nov) Instruct (2411) | 56.3 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-llm-language-compassbench-2501 · How It Works · Data refreshed daily, snapshot 2026-09-19.