Hermes 3 - Llama-3.1 70B

Creator

Nous Research

Released

2024-08-15

Intelligence

10.6

Artificial Analysis Index

Coding

Artificial Analysis Index

In $/1M

$0.30

input tokens

Out $/1M

$0.30

output tokens

Blended $/1M

$0.30

3:1 blended

Speed

32

tokens / sec

Benchmark breakdown

Independent evaluation scores, normalized to 0–100.

GPQA Diamond
40
Humanity’s Last Exam
4
MMLU-Pro
57
SciCode
23
LiveCodeBench
19
MATH-500
54
AIME 2025
τ²-Bench (agentic)
Terminal-Bench Hard
IFBench
Via Artificial Analysis