benchmark
.
space
benchmarks
rankings
compare
voices
transcripts
articles
head to head
Claude Opus 4.6
vs
GPT-5.2
5 shared benchmarks
2
wins
1
ties
2
wins
100.0%
AIME
100.0%
91.3%
GPQA Diamond
92.4%
53.1%
Humanity's Last Exam
27.8%
93.0%
MRCR v2
98.0%
81.42%
SWE-bench Verified
80.0%