← All comparisons

Mistral Small 4 (Reasoning) vs Claude Opus 4.5 (Non-reasoning)

Mistral vs Anthropic — side-by-side benchmark comparison

Mistral Small 4 (Reasoning)Claude Opus 4.5 (Non-reasoning)
Intelligence Index27.843.1
Coding Index24.342.9
Math Index62.7
Output speed (tok/s)172.957.0
Blended price ($/1M)$0.26$10.94
Time to first token (s)0.46s0.93s
aime
aime 2562.7%
artificial analysis coding index24.3042.90
artificial analysis intelligence index27.8043.10
artificial analysis math index62.70
gpqa76.9%81.0%
hle9.5%12.9%
ifbench48.2%43.0%
lcr44.7%65.3%
livecodebench73.8%
math 500
mmlu pro88.9%
scicode38.0%47.0%
tau241.2%86.3%
terminalbench hard17.4%40.9%

Benchmark data from Artificial Analysis.