← All comparisons

Devstral Small 2 vs Claude 4.1 Opus (Reasoning)

Mistral vs Anthropic — side-by-side benchmark comparison

Devstral Small 2Claude 4.1 Opus (Reasoning)
Intelligence Index19.542.0
Coding Index20.736.5
Math Index34.380.3
Output speed (tok/s)75.244.5
Blended price ($/1M)$0.00$32.81
Time to first token (s)0.73s8.55s
aime
aime 2534.3%80.3%
artificial analysis coding index20.7036.50
artificial analysis intelligence index19.5042.00
artificial analysis math index34.3080.30
gpqa53.2%80.9%
hle3.4%11.9%
ifbench31.2%55.4%
lcr24.0%66.3%
livecodebench34.8%65.4%
math 500
mmlu pro67.8%88.0%
scicode28.8%40.9%
tau223.4%71.4%
terminalbench hard16.7%34.3%

Benchmark data from Artificial Analysis.