← All comparisons

Claude 4.1 Opus (Reasoning) vs Kimi K2.5 (Non-reasoning)

Anthropic vs Kimi — side-by-side benchmark comparison

Claude 4.1 Opus (Reasoning)Kimi K2.5 (Non-reasoning)
Intelligence Index42.037.3
Coding Index36.525.8
Math Index80.3
Output speed (tok/s)44.533.5
Blended price ($/1M)$32.81$1.20
Time to first token (s)8.55s1.23s
aime
aime 2580.3%
artificial analysis coding index36.5025.80
artificial analysis intelligence index42.0037.30
artificial analysis math index80.30
gpqa80.9%78.9%
hle11.9%12.3%
ifbench55.4%43.7%
lcr66.3%59.0%
livecodebench65.4%
math 500
mmlu pro88.0%
scicode40.9%39.6%
tau271.4%81.3%
terminalbench hard34.3%18.9%

Benchmark data from Artificial Analysis.