← All comparisons

Mistral Small 4 (Non-reasoning) vs Claude 3.5 Sonnet (Oct '24)

Mistral vs Anthropic — side-by-side benchmark comparison

Mistral Small 4 (Non-reasoning)Claude 3.5 Sonnet (Oct '24)
Intelligence Index18.615.9
Coding Index16.430.2
Math Index
Output speed (tok/s)157.60.0
Blended price ($/1M)$0.26$6.56
Time to first token (s)0.49s0.00s
aime15.7%
aime 25
artificial analysis coding index16.4030.20
artificial analysis intelligence index18.6015.90
artificial analysis math index
gpqa57.1%59.9%
hle3.7%3.9%
ifbench32.8%
lcr21.3%
livecodebench38.1%
math 50077.1%
mmlu pro77.2%
scicode28.1%36.6%
tau218.4%
terminalbench hard10.6%

Benchmark data from Artificial Analysis.