← All comparisons

Hermes 4 - Llama-3.1 405B (Non-reasoning) vs Qwen3 Omni 30B A3B (Reasoning)

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

Hermes 4 - Llama-3.1 405B (Non-reasoning)Qwen3 Omni 30B A3B (Reasoning)
Intelligence Index17.615.6
Coding Index18.112.7
Math Index15.374.0
Output speed (tok/s)40.8100.1
Blended price ($/1M)$1.50$0.43
Time to first token (s)0.73s0.97s
aime
aime 2515.3%74.0%
artificial analysis coding index18.1012.70
artificial analysis intelligence index17.6015.60
artificial analysis math index15.3074.00
gpqa53.6%72.6%
hle4.2%7.3%
ifbench34.8%43.4%
lcr20.0%0.0%
livecodebench54.6%67.9%
math 500
mmlu pro72.9%79.2%
scicode34.6%30.6%
tau226.6%21.3%
terminalbench hard9.8%3.8%

Benchmark data from Artificial Analysis.