OpenCompass Language - Creation - English (CompassBench 2407): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 36 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude 3.5 Sonnet (20240620) | 55.9 |
| 2 | GPT-4o (2024-05-13) | 52.2 |
| 3 | Mistral Large 2 (Jul) | 51.6 |
| 4 | Mistral Large | 50.7 |
| 5 | GPT-4o Mini (2024-07-18) | 50.2 |
| 6 | Qwen 1.5 110B Chat | 48.9 |
| 7 | Yi 1.5 34B Chat | 46.4 |
| 8 | Yi Large | 45.2 |
| 9 | GLM-4 (0520) | 45.2 |
| 10 | Gemma 2 27B (IT) | 43.2 |
| 11 | DeepSeek V2 Chat (0628) | 42.6 |
| 12 | Yi-1.5-9B Chat | 41.2 |
| 13 | Llama 3.1 70B Instruct | 41 |
| 14 | GLM-4 9B Chat | 41 |
| 15 | Mixtral 8x7B Instruct (v0.1) | 40.5 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-language-creation-english-compassbench-2407 · How It Works · Data refreshed daily, snapshot 2026-09-19.