OpenCompass Math - College: leaderboard
OpenCompass math subscore for college-level math tasks.
Metric: Score (%). Source: rank.opencompass.org.cn. Saturation forecast: Not forecast. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.7 Flash | 85.6 | #23 |
| 2 | Grok 4.6 (High) | 83.8 | #26 (Grok 4.6) |
| 3 | Kimi K2.6 (High) | 82.7 | #104 (Kimi K2.6) |
| 4 | Hy3-preview (High) | 81 | #157 (Hy3-preview) |
| 5 | DeepSeek V4 Pro | 79.9 | #102 |
| 6 | Claude Opus 5 (High) | 79.9 | #12 (Claude Opus 5) |
| 7 | Qwen 3.8 Max | 79.2 | #35 |
| 8 | Kimi K3 (High) | 79.2 | #30 (Kimi K3) |
| 9 | GLM-5.3 | 77.1 | #39 |
| 10 | DeepSeek V4 Flash | 75 | #137 |
| 11 | Claude Opus 4.8 (High) | 74.6 | #37 (Claude Opus 4.8) |
| 12 | MiniMax-M3 | 74.3 | #136 |
| 13 | GPT-5.6 Sol (High) | 74.3 | #19 (GPT-5.6 Sol) |
| 14 | GPT-5.5 (High) | 73.9 | #31 (GPT-5.5) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=opencompass-math-college · How It Works · Data refreshed daily, snapshot 2026-10-20.