OpenCompass LLM - Agent (CompassBench 2404): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 31 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4 Turbo (Preview) | 82 |
| 2 | Qwen-72B-Chat | 58.1 |
| 3 | Qwen-14B-Chat | 54.9 |
| 4 | Yi 34B (Chat) | 53.6 |
| 5 | Qwen-7B Chat | 48.5 |
| 6 | OrionStar-Yi-34B-Chat | 42.3 |
| 7 | Yi 6B (Chat) | 40.9 |
| 8 | deepseek-llm-7B-chat | 40.5 |
| 9 | Llama 2 70B Chat | 38.6 |
| 10 | Baichuan2-7B-Chat | 37.7 |
| 11 | Mistral 7B Instruct (v0.2) | 37.3 |
| 12 | WizardLM-70B-V1.0 | 35.6 |
| 13 | WizardLM-13B-V1.2 | 34.9 |
| 14 | Baichuan2-13B-Chat | 34.3 |
| 15 | Llama 2 13B Chat | 32.5 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-llm-agent-compassbench-2404 · How It Works · Data refreshed daily, snapshot 2026-09-19.