OpenCompass Subjective - Code - English (CompassBench 2405): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 39 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o (2024-05-13) | 58.8 |
| 2 | GPT-4 Turbo | 46.7 |
| 3 | Llama 3 70B Instruct | 46.5 |
| 4 | Qwen 1.5 110B Chat | 41.9 |
| 5 | GLM-4 | 41.4 |
| 6 | Yi Large | 41 |
| 7 | Mistral Large | 40.6 |
| 8 | Claude 3 Opus | 37.8 |
| 9 | Command-R+ | 37.6 |
| 10 | Qwen 1.5 72B Chat | 37 |
| 11 | DeepSeek V2 Chat | 35.7 |
| 12 | Qwen 1.5 14B Chat | 35.1 |
| 13 | Moonshot v1 8K | 34.7 |
| 14 | Qwen 1.5 32B Chat | 32.8 |
| 15 | internlm2-chat-20B | 31.8 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-subjective-code-english-compassbench-2405 · How It Works · Data refreshed daily, snapshot 2026-09-19.