OpenCompass LLM - Reasoning: leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 5 (High) | 68.4 |
| 2 | GPT-5.5 (High) | 66.9 |
| 3 | GPT-5.6 Sol (High) | 66.8 |
| 4 | Qwen 3.8 Max | 65.2 |
| 5 | Claude Opus 4.8 (High) | 64.1 |
| 6 | Kimi K3 (High) | 63.9 |
| 7 | Gemini 3.7 Flash | 62.1 |
| 8 | GLM-5.3 | 62 |
| 9 | DeepSeek V4 Flash | 61 |
| 10 | Grok 4.6 (High) | 60 |
| 11 | DeepSeek V4 Pro | 55.5 |
| 12 | Hy3-preview (High) | 54.6 |
| 13 | MiniMax-M3 | 54 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-llm-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-05.