OpenCompass LLM - Code - Chinese (CompassBench 2405): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 39 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o (2024-05-13) | 71.2 |
| 2 | GPT-4 Turbo | 66.6 |
| 3 | DeepSeek V2 Chat | 64.5 |
| 4 | Yi Large | 59.7 |
| 5 | Claude 3 Opus | 56.9 |
| 6 | Llama 3 70B Instruct | 56 |
| 7 | GPT-3.5 Turbo | 52.9 |
| 8 | Mixtral 8x22B Instruct (v0.1) | 52.3 |
| 9 | Yi-1.5-9B Chat | 50.9 |
| 10 | GLM-4 | 50.4 |
| 11 | Qwen 1.5 110B Chat | 49.9 |
| 12 | Moonshot v1 8K | 49.7 |
| 13 | Yi 1.5 34B Chat | 47.3 |
| 14 | Mistral Large | 43.3 |
| 15 | Yi-1.5-6B-Chat | 42.9 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-llm-code-chinese-compassbench-2405 · How It Works · Data refreshed daily, snapshot 2026-09-19.