Hermes 4 - Llama-3.1 70B
Non-reasoning
Creator
Nous Research
Released
2025-08-27
Intelligence
12.6
Artificial Analysis Index
Coding
9.2
Artificial Analysis Index
In $/1M
$0.13
input tokens
Out $/1M
$0.40
output tokens
Blended $/1M
$0.20
3:1 blended
Speed
88
tokens / sec
Benchmark breakdown
Independent evaluation scores, normalized to 0–100.
GPQA Diamond
49
Humanity’s Last Exam
4
MMLU-Pro
66
SciCode
28
LiveCodeBench
27
MATH-500—
AIME 2025
11
τ²-Bench (agentic)
22
Terminal-Bench Hard
0
IFBench
29