Benchmark

Global-MMLU

A comprehensive multilingual benchmark covering 42 languages that addresses cultural and linguistic biases in evaluation, with improved translation…

Modality
text
Categories
reasoning, language, general
Openness
unknown
Source
llm_stats
Reported scores
5

Reported scores

llm_stats

ModelOrganizationReported valueReportedEvidence
Gemma 3n E2B InstructedGoogle0.5512025-06-26evidence
Gemma 3n E2B Instructed LiteRT (Preview)Google0.5512025-05-20evidence
Gemma 3n E4B InstructedGoogle0.6032025-06-26evidence
Gemma 3n E4B Instructed LiteRT PreviewGoogle0.6032025-05-20evidence
MiMo-V2.5-ProXiaomi0.8362026-04-27evidence

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.