OpenCompass Knowledge - Humanities - English (CompassBench 2501): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 40 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek R1 | 89 |
| 2 | Gemini 2.0 Pro (Preview 02-05) | 89 |
| 3 | DeepSeek V3 | 87 |
| 4 | GLM-4 Plus | 87 |
| 5 | Doubao-1.5-pro-32k-250115 | 87 |
| 6 | Claude 3.5 Sonnet (20241022) | 86 |
| 7 | O1 (2024-12-17) | 85 |
| 8 | MiniMax-Text-01 | 85 |
| 9 | GPT-4o (2024-11-20) | 83 |
| 10 | Qwen 2.5 Max | 83 |
| 11 | Gemini 2.0 Flash (001) | 82 |
| 12 | Llama 3.1 70B Instruct | 81 |
| 13 | O3 Mini (2025-01-31) | 80 |
| 14 | Yi Lightning | 80 |
| 15 | Qwen 2.5 72B Instruct | 79 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-knowledge-humanities-english-compassbench-2501 · How It Works · Data refreshed daily, snapshot 2026-09-19.