Llama 3 Instruct 70B

Last updated: Recently

Model Type

Text

Max Context

Max Output

Parameter Scale

Llama 3 Instruct 70B by Meta: Text model; TTFT 0.682s, 46.2 tok/s.

Input Modality

Output Modality

Inference Speed

46.215 tokens/s

Latest Release Date

4/18/2024

SDK Ecosystem

artificial-analysismanufactureraa-bootstrap

Model Overview Fields

name

Llama 3 Instruct 70B

Release date

4/18/2024

Performance

46.215 tokens/s

model_creator.name

Meta

AA Evaluation Scores

AA Intelligence Index

8.9

AA Coding Index

6.8

AA Math Index

MMLU Pro

0.574

GPQA

0.379

HLE

0.044

LiveCodeBench

0.198

SciCode

0.189

Math-500

0.483

AIME

AIME 2025

IFBench

0.371

LCR

TerminalBench Hard

0.008

TAU2

Output Speed (tokens/s)

46.215 tok/s

TTFT (s)

0.682s

First Answer Token (s)

0.682s