← All comparisons

Hermes 4 - Llama-3.1 70B (Non-reasoning) vs Kimi K2.5 (Non-reasoning)

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

Hermes 4 - Llama-3.1 70B (Non-reasoning)Kimi K2.5 (Non-reasoning)
Intelligence Index12.637.3
Coding Index9.225.8
Math Index11.3
Output speed (tok/s)94.333.5
Blended price ($/1M)$0.20$1.20
Time to first token (s)0.61s1.23s
aime
aime 2511.3%
artificial analysis coding index9.2025.80
artificial analysis intelligence index12.6037.30
artificial analysis math index11.30
gpqa49.1%78.9%
hle3.6%12.3%
ifbench29.0%43.7%
lcr2.0%59.0%
livecodebench26.9%
math 500
mmlu pro66.4%
scicode27.7%39.6%
tau221.6%81.3%
terminalbench hard0.0%18.9%

Benchmark data from Artificial Analysis.