← All comparisons

Hermes 4 - Llama-3.1 70B (Reasoning) vs Qwen3.6 35B A3B (Reasoning)

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

Hermes 4 - Llama-3.1 70B (Reasoning)Qwen3.6 35B A3B (Reasoning)
Intelligence Index16.043.5
Coding Index14.435.2
Math Index68.7
Output speed (tok/s)92.8159.1
Blended price ($/1M)$0.20$0.56
Time to first token (s)0.64s1.50s
aime
aime 2568.7%
artificial analysis coding index14.4035.20
artificial analysis intelligence index16.0043.50
artificial analysis math index68.70
gpqa69.9%84.1%
hle7.9%20.2%
ifbench31.3%64.4%
lcr6.7%63.7%
livecodebench65.3%
math 500
mmlu pro81.1%
scicode34.1%35.8%
tau222.5%95.3%
terminalbench hard4.5%34.8%

Benchmark data from Artificial Analysis.