Llama Guard 4 is a multimodal content safety classifier derived from Llama 4 Scout and fine-tuned for both prompt and response classification. It acts as an LLM: it generates text that indicates whether a given prompt or response is safe or unsafe and, if unsafe, lists the content categories violated. It is aligned with the standardized MLCommons hazards taxonomy and moderates English and multiple other supported languages, including mixed text-and-image prompts with multiple images. It is integrated into the Llama Moderations API for safety classification of text and images.
$0.072market price $0.144
$0.072market price $0.144
Throughput is how fast the model writes (tokens per second, higher is better). Time to first token is how long before the answer starts (lower is better).
Share of requests served without a server error. Client errors (4xx) do not count.
Not enough throughput data
Not enough TTFT data
Works with the OpenAI SDK pointed at our base URL. Every parameter is in the API reference.
Tokens processed per day on NagaAI.
Llama Guard 4 is a multimodal content safety classifier derived from Llama 4 Scout and fine-tuned for both prompt and response classification. It acts as an LLM: it generates text that indicates whether a given prompt or response is safe or unsafe and, if unsafe, lists the content categories violated.
On NagaAI, Llama Guard 4 12B costs $0.072 per 1M input tokens and $0.072 per 1M output tokens. You pay only for what you use, with no subscription.
Llama Guard 4 12B reads up to 163.8k tokens in one request, counting the prompt and the answer together.
Yes. Llama Guard 4 12B can call your functions as tools and can answer in a JSON schema you give it.
It takes text and images as input and returns text.
Create an API key, copy the model ID from the top of this page and send requests with the OpenAI or Anthropic SDK pointed at NagaAI. The API section on this page has a ready example.