Claude Opus 4.6 vs Grok 4
6 shared benchmarks
4
wins
0
ties
2
wins
100.0%
AIME
94.0%
72.0%
Aider Polyglot
79.6%
91.3%
GPQA Diamond
88.0%
53.1%
Humanity's Last Exam
24.0%
82.0%
MMLU Pro
87.0%
81.42%
SWE-bench Verified
58.6%