OpenCompass Language - Traditional Cultural Understanding - Chinese (CompassBench 2404): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 31 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | OrionStar-Yi-34B-Chat | 65.3 |
| 2 | Yi 34B (Chat) | 59.3 |
| 3 | Qwen-72B-Chat | 58 |
| 4 | Qwen-14B-Chat | 54 |
| 5 | Yi 6B (Chat) | 47.7 |
| 6 | Qwen-7B Chat | 42.3 |
| 7 | GPT-4 Turbo (Preview) | 36 |
| 8 | Baichuan2-13B-Chat | 32.3 |
| 9 | Baichuan2-7B-Chat | 32 |
| 10 | deepseek-llm-7B-chat | 26.7 |
| 11 | WizardLM-70B-V1.0 | 12.7 |
| 12 | WizardLM-13B-V1.2 | 7 |
| 13 | Llama 2 70B Chat | 6.3 |
| 14 | Mistral 7B Instruct (v0.2) | 5.3 |
| 15 | Llama 2 7B Chat | 4.3 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-language-traditional-cultural-understanding-chinese-compassbench-2404 · How It Works · Data refreshed daily, snapshot 2026-09-19.