benchmark
.
space
benchmarks
rankings
compare
voices
transcripts
papers
articles
head to head
Claude Opus 4.6
vs
Claude Opus 4.7
15 shared benchmarks
1
wins
0
ties
14
wins
72.0%
Aider Polyglot
79.0%
90.2%
BigLaw Bench
90.9%
84.0%
BrowseComp
79.3%
61.5%
CharXiv
89.0%
66.6%
CyberGym
73.8%
91.3%
GPQA Diamond
94.2%
76.0%
LiveCodeBench
83.0%
82.0%
MMLU Pro
87.0%
91.1%
MMMLU
92.0%
72.7%
OSWorld
78.0%
77.8%
SWE-bench Multilingual
85.7%
27.1%
SWE-bench Multimodal
35.0%
53.4%
SWE-bench Pro
64.3%
81.42%
SWE-bench Verified
87.6%
65.4%
Terminal-Bench 2.0
77.0%