xAI
Grok 4
10 benchmarks
AIME
#30 of 117
94.0%
GPQA Diamond
#37 of 118
88.0%
GPQA Diamond
#37 of 118
87.5%
MMLU Pro
#8 of 55
87.0%
Aider Polyglot
#6 of 12
79.6%
LiveCodeBench
#24 of 53
79.0%
ARC-AGI
#1 of 1
66.6%
SWE-bench Verified
#73 of 86
58.6%
Humanity's Last Exam
#59 of 72
24.0%
ARC-AGI 2
#11 of 11
15.9%
benchmark.space canonical public ledger of AI capability