Llama 3 Instruct 8B

Last updated: Recently

Model Type

Text

Max Context

Max Output

Parameter Scale

Llama 3 Instruct 8B by Meta: Text model; TTFT 0.498s, 82.3 tok/s.

Input Modality

Output Modality

Inference Speed

82.251 tokens/s

Latest Release Date

4/18/2024

SDK Ecosystem

artificial-analysismanufactureraa-bootstrap

Model Overview Fields

name

Llama 3 Instruct 8B

Release date

4/18/2024

Performance

82.251 tokens/s

model_creator.name

Meta

AA Evaluation Scores

AA Intelligence Index

6.4

AA Coding Index

AA Math Index

MMLU Pro

0.405

GPQA

0.296

HLE

0.051

LiveCodeBench

0.096

SciCode

0.119

Math-500

0.499

AIME

AIME 2025

IFBench

0.246

LCR

TerminalBench Hard

TAU2

Output Speed (tokens/s)

82.251 tok/s

TTFT (s)

0.498s

First Answer Token (s)

0.498s