Skip to content
Bencher

TestsBenchmarks · Math

AIME 2025

2025 American Invitational Mathematics Examination.

% solvedHigher is betterToo easy now: the top models all score near perfectSaturatedThe test's websiteOfficial page ↗

The best 2 · 2 models tested

Top 2 · best setting per model · 2 models, 2 results (0 independent)

Measured by independent testersIndependent Reported by the makerVendor-reported
AIME 2025 leaderboard
Thinking levelSetting
1Mistral Medium 3.5Mistral AI86.3%Highmaximum reasoning settings (reasoning_effort=high)Maker's own figureVendor-reportedMistral AI ↗—22 May 2026
2Mistral Small 4Mistral AI83.8%HighMistral Small 4 - HighMaker's own figureVendor-reportedMistral AI ↗—16 Mar 2026