# Llama Guard 4 12B

> Llama Guard 4 is a 12B natively multimodal safety classifier for content safety classification on LLM inputs and responses. It acts as an LLM itself: it generates text indicating whether a prompt or response is safe or unsafe and, if unsafe, lists the violated categories (MLCommons hazards taxonomy S1–S14). Use it as a guardrail/judge hop in front of or behind chat models.

**ID**: `meta/llama-guard-4-12b`  
**Creator**: meta  
**Category**: multimodal  
**Context**: 164K tokens  
**Input**: $0.18/M  
**Output**: $0.18/M  
**Released**: 2025-04-29  
**Web page**: https://anyrouter.dev/model/meta/llama-guard-4-12b

**Input modalities**: text, image  
**Output modalities**: text  

**Capabilities**: chat, streaming, vision

**Supported parameters**: max_tokens, temperature, top_p, stop

**Aliases**: `meta-llama/llama-guard-4-12b`

## Usage

**Endpoint**: `POST https://anyrouter.dev/api/v1/chat/completions`

```bash
curl https://anyrouter.dev/api/v1/chat/completions \
  -H "Authorization: Bearer $ANYROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "meta/llama-guard-4-12b",
  "messages": [
    {
      "role": "user",
      "content": "Say hi in 3 words."
    }
  ]
}'
```

## Providers

| Provider | Input | Output |
| --- | --- | --- |
| deepinfra-byok | $0 | $0 |
| openrouter-byok | $0 | $0 |

## Source

- Upstream docs: https://deepinfra.com/meta-llama/Llama-Guard-4-12B