← All comparisons

Gemini 3.1 Flash-Lite vs Hermes 3 - Llama-3.1 70B

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

Gemini 3.1 Flash-LiteHermes 3 - Llama-3.1 70B
Intelligence Index33.510.6
Coding Index30.1
Math Index
Output speed (tok/s)304.933.2
Blended price ($/1M)$0.56$0.30
Time to first token (s)5.00s0.38s
aime2.3%
aime 25
artificial analysis coding index30.10
artificial analysis intelligence index33.5010.60
artificial analysis math index
gpqa82.2%40.1%
hle16.2%4.1%
ifbench77.2%
lcr65.3%
livecodebench18.8%
math 50053.8%
mmlu pro57.1%
scicode41.9%23.1%
tau231.3%
terminalbench hard24.2%

Benchmark data from Artificial Analysis.