Benchmarks

GPQA ยท Mainreasoning%225
MMLU-Promath%127
AIME 2025math%115
MMLUmath%105
SWE-Bench Verifiedcoding%101
Humanity's Last Exammath%86
MATHmath%75
LiveCodeBenchcoding%74
HumanEvalcoding%72
MMMUreasoning%66
IFEvalgeneral%64
MMMU-Proreasoning%56
AIME 2024math%54
LiveCodeBench v6reasoning%50
MMMLUmath%49
BrowseCompreasoning%48
GSM8kmath%48
MMLU-Reduxmath%47
SimpleQAreasoning%46
Terminal-Bench 2.0coding%46