OpenCompass Math - Competition (CompassBench 2601): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 26 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.2 (High) | 70 |
| 2 | Qwen 3 Max (2026-01-23) | 70 |
| 3 | Kimi K2.5 | 69.2 |
| 4 | Gemini 3 Pro (Preview) | 69 |
| 5 | Kimi K2 (Thinking) | 66.3 |
| 6 | Claude Opus 4.5 (Thinking) | 65 |
| 7 | GPT-OSS-120B (High) | 64.5 |
| 8 | DeepSeek V3.2 | 64.2 |
| 9 | GLM-4.7 | 64 |
| 10 | Tencent HY 2.0 Think | 62.9 |
| 11 | Ring-1T | 61.8 |
| 12 | Grok 4.1 | 61 |
| 13 | Claude Sonnet 4.5 (Thinking) | 60.3 |
| 14 | Qwen 3 235B A22B 2507 (Thinking) | 58.7 |
| 15 | Qwen 3 Next 80B A3B (Thinking) | 58.4 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-math-competition-compassbench-2601 · How It Works · Data refreshed daily, snapshot 2026-09-19.