MathArena - CMIMC 2025 — leaderboard
Metric: Accuracy (%). Source: matharena.ai. 36 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V3.2 Speciale | 94.38 |
| 2 | Step 3.5 Flash | 93.75 |
| 3 | GLM-5 | 92.5 |
| 4 | Kimi K2 (Thinking) | 91.88 |
| 5 | GPT-5.1 (High) | 91.88 |
| 6 | Kimi K2.5 (Thinking) | 91.25 |
| 7 | GPT-5.2 (High) | 91.25 |
| 8 | Gemini 3 Flash | 90.62 |
| 9 | Gemini 3 Pro (Preview) | 90 |
| 10 | GPT-5 (High) | 90 |
| 11 | GLM-4.6 | 88.75 |
| 12 | GPT-OSS-120B (High) | 85.62 |
| 13 | Grok 4 Fast (Reasoning) | 85.62 |
| 14 | Grok 4.1 Fast (Reasoning) | 84.38 |
| 15 | GPT-5 Mini (High) | 84.38 |
Interactive version: theaggregate.ai/benchmark?slug=matharena-cmimc-2025 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.