Benchmark

AIME 2025/2026 (MathArena)

Measures: Each year's American Invitational Mathematics Examination problems, run by MathArena soon after the contest; scored as accuracy averaged over repeated runs with a 95% CI.

Maintainer
SRI Lab, ETH Zurich and INSAIT (MathArena)
Introduced
2025
Status
Superseded
Saturation
Saturated
Typical provenance
Mixed
In scoring set
No

PaperOfficial leaderboard

Leaderboard

Independent and third-party results only.

No independent or third-party results yet. Show all results

AIME 2025/2026 (MathArena): leaderboard and what it measures · supermarket.dev