OpenCompass Code - Comprehensive — leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 23 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 (High) | 63.7 |
| 2 | GPT-5.4 (High) | 63.4 |
| 3 | Kimi K2.6 | 62.4 |
| 4 | Qwen 3.6 Max Preview | 61.6 |
| 5 | Gemini 3.1 Pro (Preview) | 60.2 |
| 6 | DeepSeek V4 Pro | 59.4 |
| 7 | Claude Sonnet 4.6 (High) | 58.1 |
| 8 | Qwen 3.5 397B A17B | 57.4 |
| 9 | Gemma 4 31B (IT) | 56.8 |
| 10 | DeepSeek V4 Flash | 56.4 |
| 11 | MiniMax-M2.7 | 56.2 |
| 12 | GLM-5.1 | 55.7 |
| 13 | Step 3.5 Flash | 54.9 |
| 14 | GPT-5.4 Mini (High) | 52.9 |
| 15 | Gemini 3.1 Flash Lite (Preview) | 52.8 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-code-comprehensive · How the rankings work · Data refreshed daily, snapshot 2026-07-22.