Benchmark

AIME

American Invitational Mathematics Examination (AIME) benchmark for evaluating mathematical reasoning capabilities of large language models. Contains 30…

Modality
text
Categories
math, reasoning
Openness
unknown
Source
llm_stats
Reported scores
2

Reported scores

llm_stats

ModelOrganizationReported valueReportedEvidence
MiMo-V2.5-ProXiaomi0.3732026-04-27evidence
Phi 4 Mini ReasoningMicrosoft0.5752025-04-30evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.