Skip to main content
llm.info

Llama Guard

Open Source
Library

Open safety classifier for LLM I/O

About

Llama Guard is Meta's open-weight safeguard model that classifies prompts and responses for safety across a taxonomy of harms, used to moderate inputs and outputs in LLM and agent pipelines.

Compatibility

Supported Languages

python

Details

Category
Library
License
Llama Community License