OpenCompass Math - Competition: leaderboard
OpenCompass math subscore for competition-style math tasks.
Metric: Score (%). Source: rank.opencompass.org.cn. Saturation forecast: Around July 2028. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5 (High) | 74.7 | #12 (Claude Opus 5) |
| 2 | Qwen 3.8 Max | 70.8 | #35 |
| 3 | GPT-5.5 (High) | 68.9 | #31 (GPT-5.5) |
| 4 | Kimi K3 (High) | 68.9 | #30 (Kimi K3) |
| 5 | GPT-5.6 Sol (High) | 68.6 | #19 (GPT-5.6 Sol) |
| 6 | Kimi K2.6 (High) | 68.6 | #104 (Kimi K2.6) |
| 7 | Claude Opus 4.8 (High) | 68.3 | #37 (Claude Opus 4.8) |
| 8 | Gemini 3.7 Flash | 67.9 | #23 |
| 9 | DeepSeek V4 Flash | 65.7 | #137 |
| 10 | Hy3-preview (High) | 63.5 | #157 (Hy3-preview) |
| 11 | DeepSeek V4 Pro | 62.5 | #102 |
| 12 | Grok 4.6 (High) | 62.2 | #26 (Grok 4.6) |
| 13 | GLM-5.3 | 57.4 | #39 |
| 14 | MiniMax-M3 | 51.6 | #136 |
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-competition · How It Works · Data refreshed daily, snapshot 2026-10-20.