Kimi K2.5 vs Claude Mythos Preview
7 shared benchmarks
0
wins
0
ties
7
wins
60.6%
BrowseComp
86.9%
87.6%
GPQA Diamond
94.6%
24.37%
Humanity's Last Exam
56.8%
63.3%
OSWorld
79.6%
50.7%
SWE-bench Pro
77.8%
76.8%
SWE-bench Verified
93.9%
50.8%
Terminal-Bench 2.0
82.0%