Llama

Llama Guard 4 12B

meta-llama/Llama-Guard-4-12B

Create Deployment Open Playground

Overview

Llama Guard 4 is a 12B dense multimodal safety model released by Meta in April 2025, pruned from the Llama 4 Scout model. It is designed to detect inappropriate content in both images and text used as input or generated as output by models. Llama Guard 4 can evaluate text-only and image+text inputs, supports multiple languages, and classifies hazards based on the MLCommons taxonomy and code interpreter abuse, enabling flexible moderation pipelines.

Tags

CentML Optimized

Chat

Dedicated

LLM

Serverless

API

Select a language: 
curl -X POST "https://api.centml.com/openai/v1/chat/completions" \
  -H "Authorization: Bearer *******************************************" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "meta-llama/Llama-Guard-4-12B",
    "messages": [{ "role": "system", "content": "You are a helpful assistant." }],
    "stream": false
  }'
from openai import OpenAI

client = OpenAI(
	api_key="*******************************************",
	base_url="https://api.centml.com/openai/v1"
)

completion = client.chat.completions.create(
    model="meta-llama/Llama-Guard-4-12B",
    messages=[{ "role": "system", "content": "You are a helpful assistant." }],
    stream=False,
)

print(completion.choices[0].message)
import OpenAI from "openai";

const client = new OpenAI(
	api_key="*******************************************",
	base_url="https://api.centml.com/openai/v1"
)

async function main() {
  const completion = await client.chat.completions.create({
    model: "meta-llama/Llama-Guard-4-12B",
    messages: [{ "role": "system", "content": "You are a helpful assistant." }],
    stream: false,
  });

  console.log(completion.choices[0])
}

main()