Benchmark · KnowledgeSettled

MMMLU

Updated 2026-04-07
Models tested
4
Top score
92.7
Claude Mythos Preview
Median
92.7
min 92.7
Top-5 spread
σ 0.0
Settled
MMMLU \u00B7 TOP 40255075100#1Claude Mythos Preview92.7#2Gemini 3.1 ProVERIFIED92.6#3Claude Opus 4.7VERIFIED91.5#4Claude Opus 4.6VERIFIED91.1benchgecko.ai/benchmark/mmmlu

Same category · related evaluations