Llama 3.2 Instruct 3B

Last updated: Recently

Model Type

Text

Max Context

Max Output

Parameter Scale

Llama 3.2 Instruct 3B by Meta: Text model; TTFT 0.657s, 52.0 tok/s.

Input Modality

Output Modality

Inference Speed

51.984 tokens/s

Latest Release Date

9/25/2024

SDK Ecosystem

artificial-analysismanufactureraa-bootstrap

Model Overview Fields

name

Llama 3.2 Instruct 3B

Release date

9/25/2024

Performance

51.984 tokens/s

model_creator.name

Meta

AA Evaluation Scores

AA Intelligence Index

9.7

AA Coding Index

AA Math Index

3.3

MMLU Pro

0.347

GPQA

0.255

HLE

0.052

LiveCodeBench

0.083

SciCode

0.052

Math-500

0.489

AIME

0.067

AIME 2025

0.033

IFBench

0.262

LCR

0.02

TerminalBench Hard

TAU2

0.211

Output Speed (tokens/s)

51.984 tok/s

TTFT (s)

0.657s

First Answer Token (s)

0.657s