Llama Guard 4 12B
JSON →A 12-billion-parameter safety classifier from Meta designed to detect harmful content and enforce content policies in LLM inputs and outputs.
Specs
context window 8K tokens
max output 8K tokens
input price $0.2 / 1M tokens
output price $0.2 / 1M tokens
Capabilities
json-modestreaming
Dates
releasedApr 2025
knowledge cutoffJun 2024
Resources
homepagellama.meta.com/ ↗
API
full doc /v1/models/llama-guard-4-12b