OpenCompass Code - HumanEval-X (CompassBench 2405): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 39 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o (2024-05-13) | 69.5 |
| 2 | Claude 3 Opus | 68.7 |
| 3 | GPT-4 Turbo | 66.6 |
| 4 | DeepSeek V2 Chat | 60.1 |
| 5 | Mixtral 8x22B Instruct (v0.1) | 57 |
| 6 | GLM-4 | 54.5 |
| 7 | Yi 1.5 34B Chat | 54.1 |
| 8 | Yi-1.5-9B Chat | 53.9 |
| 9 | Mistral Large | 53.9 |
| 10 | Yi Large | 53 |
| 11 | Moonshot v1 8K | 51.8 |
| 12 | Qwen 1.5 110B Chat | 50.2 |
| 13 | DBRX | 47.8 |
| 14 | Qwen 1.5 32B Chat | 46.3 |
| 15 | Yi-1.5-6B-Chat | 46.2 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-code-humaneval-x-compassbench-2405 · How It Works · Data refreshed daily, snapshot 2026-09-19.