← All comparisons

Claude 4.5 Haiku (Non-reasoning) vs GPT-4o mini

Anthropic vs OpenAI — side-by-side benchmark comparison

Claude 4.5 Haiku (Non-reasoning)GPT-4o mini
Intelligence Index31.012.6
Coding Index29.6
Math Index39.014.7
Output speed (tok/s)105.869.6
Blended price ($/1M)$2.19$0.26
Time to first token (s)0.65s0.62s
aime11.7%
aime 2539.0%14.7%
artificial analysis coding index29.60
artificial analysis intelligence index31.0012.60
artificial analysis math index39.0014.70
gpqa64.6%42.6%
hle4.3%4.0%
ifbench42.0%31.0%
lcr43.7%
livecodebench51.1%23.4%
math 50078.9%
mmlu pro80.0%64.8%
scicode34.4%22.9%
tau232.5%
terminalbench hard27.3%

Benchmark data from Artificial Analysis.