Claude Opus 4.6 vs Kimi K2.5
9 shared benchmarks
7
wins
0
ties
2
wins
100.0%
AIME
96.1%
84.0%
BrowseComp
60.6%
91.3%
GPQA Diamond
87.6%
53.1%
Humanity's Last Exam
24.37%
76.0%
LiveCodeBench
85.0%
82.0%
MMLU Pro
87.1%
72.7%
OSWorld
63.3%
81.42%
SWE-bench Verified
76.8%
81.8%
Terminal-Bench 2.0
50.8%