MathArena - AIME 2025 — leaderboard

Metric: Accuracy (%). Source: matharena.ai. 61 models tracked.

Top models

#ModelScore
1GPT-5.2 (High)100
2GPT-5.2 (xHigh)100
3Human Expert100
4Step 3.5 Flash98.33
5Gemini 3 Flash97.5
6GLM-596.67
7DeepSeek V3.2 Speciale95.83
8Kimi K2.5 (Thinking)95.83
9Gemini 3 Pro (Preview)95
10GPT-5 (High)95
11DeepSeek V3.2 (Thinking)94.17
12GPT-5.1 (High)94.17
13GLM-4.593.33
14Grok 492.5
15Kimi K2 (Thinking)92.5

Interactive version: theaggregate.ai/benchmark?slug=matharena-aime-2025 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.