← All comparisons

Claude 3.5 Haiku vs Qwen3 30B A3B 2507 Instruct

Anthropic vs Alibaba — side-by-side benchmark comparison

Claude 3.5 HaikuQwen3 30B A3B 2507 Instruct
Intelligence Index18.715.0
Coding Index10.714.2
Math Index66.3
Output speed (tok/s)0.0102.1
Blended price ($/1M)$1.75$0.21
Time to first token (s)0.00s0.98s
aime3.3%72.7%
aime 2566.3%
artificial analysis coding index10.7014.20
artificial analysis intelligence index18.7015.00
artificial analysis math index66.30
gpqa40.8%65.9%
hle3.5%6.8%
ifbench42.8%33.1%
lcr23.3%22.7%
livecodebench31.4%51.5%
math 50072.1%97.5%
mmlu pro63.4%77.7%
scicode27.4%30.4%
tau224.6%10.2%
terminalbench hard2.3%6.1%

Benchmark data from Artificial Analysis.