Skip to content

General model capability

AIME 2026 / HMMT 2026

Recent competition-math reasoning.

About AIME 2026 / HMMT 2026

What it measures

Final-answer accuracy under fixed sampling and tool policy.

Scope & limitations

HF cards list CC-BY-NC-SA-4.0; clarify rights before a commercial evaluation offering. Pin year, competition and exact questions.

Evaluation environment

Mathematical reasoning model.

Source

Original benchmark

Dataset & task catalog ↗