OpenCompass Agent - T-Eval - Chinese (CompassBench 2405): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 39 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Mixtral 8x22B Instruct (v0.1) | 85.4 |
| 2 | GPT-4 Turbo | 84.9 |
| 3 | Mistral Large | 84.9 |
| 4 | Yi Large | 84.6 |
| 5 | Claude 3 Opus | 83.9 |
| 6 | Command-R+ | 83.1 |
| 7 | GPT-4o (2024-05-13) | 81.9 |
| 8 | Yi 1.5 34B Chat | 79.4 |
| 9 | Qwen 1.5 32B Chat | 75.9 |
| 10 | DeepSeek V2 Chat | 75.6 |
| 11 | Mistral 7B Instruct (v0.3) | 74.3 |
| 12 | GLM-4 | 74.2 |
| 13 | GPT-3.5 Turbo | 72.5 |
| 14 | DeepSeek-V2-Lite-Chat | 71.2 |
| 15 | Llama 3 70B Instruct | 70 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-agent-t-eval-chinese-compassbench-2405 · How It Works · Data refreshed daily, snapshot 2026-09-19.