← All comparisons

DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning) vs GPT-5 nano (high)

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

DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning)GPT-5 nano (high)
Intelligence Index7.626.8
Coding Index20.3
Math Index83.7
Output speed (tok/s)0.0173.3
Blended price ($/1M)$0.00$0.14
Time to first token (s)0.00s78.29s
aime0.0%
aime 2583.7%
artificial analysis coding index20.30
artificial analysis intelligence index7.6026.80
artificial analysis math index83.70
gpqa27.0%67.6%
hle4.3%8.2%
ifbench67.6%
lcr41.7%
livecodebench8.5%78.9%
math 50021.8%
mmlu pro36.5%78.0%
scicode9.1%36.6%
tau236.5%
terminalbench hard12.1%

Benchmark data from Artificial Analysis.