llama-3-8b-Instruct

llama38B

Directly quantized 4bit model with `bitsandbytes`.

Context
8K

Capabilities

Reasoning & Interaction
Tool Calling
Infrastructure
Streaming
Fine-tuning

Details

Model IDunsloth/llama-3-8b-Instruct
Provider Hugging Face
Creator unsloth
Familyllama-3
Licensellama3
Status—
Typechat
Input modalitiestext
Output modalitiestext
Open weightYes
Parameters8B
Architecturetransformer
Context window8K tokens
Max input—
Max output—
Reasoning tokens—
Knowledge cutoff—
Release date
Sourceofficial

Tools

Function Callingfunction_calling
Call external functions and APIs

Family Comparison: llama-3

ModelContextMax out
llama-3-70B-Instruct-abliterated8K—
llama-3-typhoon-v1.5-8b-instruct8K—
llama-3-8b-Instruct8K—

API

GET/v1/models/huggingface/unsloth/llama-3-8b-Instruct

llama-3-8b-Instruct

unsloth/llama-3-8b-Instruct

Directly quantized 4bit model with `bitsandbytes`.

Changes · 4 entries
unsloth/llama-3-8b-Instructupdate1cd659eMar 24, 2026, 04:19 AM
capabilitiesfine tuningyes
unsloth/llama-3-8b-Instructupdateb301ebeMar 23, 2026, 11:47 PM
modalitiesinput["text"]output["text"]
unsloth/llama-3-8b-Instructupdate2b100acMar 23, 2026, 05:30 AM
capabilitiestool callyes
model_typechat
parameters8
page_urlhttps://huggingface.co/unsloth/llama-3-8b-Instruct
descriptionDirectly quantized 4bit model with `bitsandbytes`.
release_date2024-04-18
context_window8192
licensellama3
taglineDirectly quantized 4bit model with `bitsandbytes`.
architecturetransformer
open_weightyes
tools["function_calling"]
unsloth/llama-3-8b-Instructcreate7f5c66fMar 21, 2026, 05:16 AM