MathArena - AIME 2025 — leaderboard
Metric: Accuracy (%). Source: matharena.ai. 61 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.2 (High) | 100 |
| 2 | GPT-5.2 (xHigh) | 100 |
| 3 | Human Expert | 100 |
| 4 | Step 3.5 Flash | 98.33 |
| 5 | Gemini 3 Flash | 97.5 |
| 6 | GLM-5 | 96.67 |
| 7 | DeepSeek V3.2 Speciale | 95.83 |
| 8 | Kimi K2.5 (Thinking) | 95.83 |
| 9 | Gemini 3 Pro (Preview) | 95 |
| 10 | GPT-5 (High) | 95 |
| 11 | DeepSeek V3.2 (Thinking) | 94.17 |
| 12 | GPT-5.1 (High) | 94.17 |
| 13 | GLM-4.5 | 93.33 |
| 14 | Grok 4 | 92.5 |
| 15 | Kimi K2 (Thinking) | 92.5 |
Interactive version: theaggregate.ai/benchmark?slug=matharena-aime-2025 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.