← All comparisons

Hermes 4 - Llama-3.1 70B (Non-reasoning) vs Qwen3 Omni 30B A3B Instruct

Nous Research vs Alibaba — side-by-side benchmark comparison

Hermes 4 - Llama-3.1 70B (Non-reasoning)Qwen3 Omni 30B A3B Instruct
Intelligence Index12.610.7
Coding Index9.27.2
Math Index11.352.3
Output speed (tok/s)94.3108.9
Blended price ($/1M)$0.20$0.43
Time to first token (s)0.61s1.04s
aime
aime 2511.3%52.3%
artificial analysis coding index9.207.20
artificial analysis intelligence index12.6010.70
artificial analysis math index11.3052.30
gpqa49.1%62.0%
hle3.6%5.1%
ifbench29.0%31.2%
lcr2.0%0.0%
livecodebench26.9%42.2%
math 500
mmlu pro66.4%72.5%
scicode27.7%18.6%
tau221.6%16.4%
terminalbench hard0.0%1.5%

Benchmark data from Artificial Analysis.