OpenCompass Subjective - Creation - Chinese (CompassBench 2405): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 39 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o (2024-05-13) | 64.2 |
| 2 | Qwen 1.5 110B Chat | 55.9 |
| 3 | Yi 1.5 34B Chat | 50.8 |
| 4 | internlm2-chat-20B | 50.4 |
| 5 | Yi Large | 49.2 |
| 6 | GLM-4 | 48.1 |
| 7 | Moonshot v1 8K | 46.9 |
| 8 | DeepSeek V2 Chat | 46.4 |
| 9 | Claude 3 Opus | 46.2 |
| 10 | Yi-1.5-9B Chat | 45.3 |
| 11 | GPT-4 Turbo | 42.5 |
| 12 | internlm2-chat-7B | 41.4 |
| 13 | Command-R+ | 37.1 |
| 14 | Yi-1.5-6B-Chat | 33 |
| 15 | Qwen 1.5 32B Chat | 29.5 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-subjective-creation-chinese-compassbench-2405 · How It Works · Data refreshed daily, snapshot 2026-09-19.