AIME 2025: leaderboard

All 30 problems from the 2025 American Invitational Mathematics Examination, testing olympiad-level mathematical reasoning.

Metric: Accuracy (%). Source: artificialanalysis.ai. Status: saturated. 178 models tracked.

Top models

#ModelScore
1Best Expert100
2GPT-5 Codex (High)98.67
3MiMo-V2-Flash (Reasoning)96.33
4Gemini 3 Pro (Preview) (High)95.67
5GLM-4.7 (Reasoning)95
6Kimi K2 (Thinking)94.67
7GPT-5 (High)94.33
8GPT-5.1 (High)94
9GPT-OSS-120B (High)93.44
10Grok 492.67
11GPT-5 (Medium)91.67
12Claude Opus 4.5 (Thinking)91.33
13Qwen 3 235B A22B 2507 (Thinking)91
14GPT-5 Mini (High)90.67
15Grok 4 Fast (Reasoning)89.67

Interactive version: theaggregate.ai/benchmark?slug=aime-2025 · How It Works · Data refreshed daily, snapshot 2026-09-05.