Llama 3.1 Instruct 70B

Creator

Meta AI (FAIR)

Released

2024-07-23

Intelligence

12.5

Artificial Analysis Index

Coding

10.9

Artificial Analysis Index

In $/1M

$0.56

input tokens

Out $/1M

$0.56

output tokens

Blended $/1M

$0.56

3:1 blended

Speed

38

tokens / sec

Benchmark breakdown

Independent evaluation scores, normalized to 0–100.

GPQA Diamond
41
Humanity’s Last Exam
5
MMLU-Pro
68
SciCode
27
LiveCodeBench
23
MATH-500
65
AIME 2025
4
τ²-Bench (agentic)
15
Terminal-Bench Hard
3
IFBench
34
Via Artificial Analysis