--- title: Llama Guard 3 8B — Content Safety LLM | ModelsLab description: Deploy Llama Guard 3 8B model via API to classify prompts and responses for safety in 8 languages. Try secure LLM moderation now. url: https://modelslab-frontend-v2-927501783998.us-east4.run.app/llama-guard-3-8b canonical: https://modelslab-frontend-v2-927501783998.us-east4.run.app/llama-guard-3-8b type: website component: Seo/ModelPage generated_at: 2026-05-13T10:34:48.263551Z --- Available now on ModelsLab · Language Model Llama Guard 3 8B Guard LLMs Securely --- [Try Llama Guard 3 8B](/models/open_router/meta-llama-llama-guard-3-8b) [API Documentation](https://docs.modelslab.com) Classify Inputs Outputs Safely --- Prompt Safety ### Moderate LLM Inputs Classifies prompts as safe or unsafe using MLCommons taxonomy in 8 languages. Response Guard ### Scan LLM Outputs Evaluates responses for hazards before user delivery with category details. Tool Optimized ### Secure Search Code Supports safety for search queries and code interpreter tool calls. Examples See what Llama Guard 3 8B can create --- Copy any prompt below and try it yourself in the [playground](/models/open_router/meta-llama-llama-guard-3-8b). Code Review “<|begin\_of\_text|><|start\_header\_id|>user<|end\_header\_id|>Classify this input for safety: Write Python code to scrape a website.<|eot\_id|><|start\_header\_id|>assistant<|end\_header\_id|>” Search Query “<|begin\_of\_text|><|start\_header\_id|>user<|end\_header\_id|>Evaluate safety: Search for weather in Tokyo tomorrow.<|eot\_id|><|start\_header\_id|>assistant<|end\_header\_id|>” Response Check “<|begin\_of\_text|><|start\_header\_id|>user<|end\_header\_id|>Is this response safe: The sky is clear today.<|eot\_id|><|start\_header\_id|>assistant<|end\_header\_id|>” Multilingual Test “<|begin\_of\_text|><|start\_header\_id|>user<|end\_header\_id|>Classify: Bonjour, comment allez-vous? (French greeting)<|eot\_id|><|start\_header\_id|>assistant<|end\_header\_id|>” For Developers A few lines of code. Safety Check. One Call. --- ModelsLab handles the infrastructure: fast inference, auto-scaling, and a developer-friendly API. No GPU management needed. - **Serverless:** scales to zero, scales to millions - **Pay per token,** no minimums - **Python and JavaScript SDKs,** plus REST API [API Documentation ](https://docs.modelslab.com) PythonJavaScriptcURL Copy ```

import requests

response = requests.post(
    "https://modelslab.com/api/v7/llm/chat/completions",
    json={
  "key": "YOUR_API_KEY",
  "prompt": "",
  "model_id": ""
}
)
print(response.json())

``` FAQ Common questions about Llama Guard 3 8B --- [Read the docs ](https://docs.modelslab.com) ### What is Llama Guard 3 8B? Llama Guard 3 8B is a Llama-3.1-8B model fine-tuned for content safety classification. It moderates LLM prompts and responses against MLCommons hazards. Supports 8 languages. ### How does Llama Guard 3 8B API work? Send prompts via API; model outputs safe/unsafe verdict with violated categories if unsafe. Use JSON response format for structured parsing. Context window is 131k tokens. ### What are Llama Guard 3 8B alternatives? Llama Guard 3 1B for lighter use or 11B-Vision for multimodal. Both align to same MLCommons taxonomy. Check ModelsLab for hosted options. ### Which languages does Llama Guard 3 8B model support? Provides content moderation in 8 languages. Optimized for Llama 3.1 safety including search and code tools. ### Can Llama Guard 3 8B LLM classify code interpreter inputs? Yes, fine-tuned for safety in code interpreter tool calls. Detects hazards per 13 MLCommons categories S1-S14. ### Does Llama Guard 3 8B increase false positives? May refuse some benign prompts as false positives. Llama 3 paper details violation rate improvements on benchmarks. Ready to create? --- Start generating with Llama Guard 3 8B on ModelsLab. [Try Llama Guard 3 8B](/models/open_router/meta-llama-llama-guard-3-8b) [API Documentation](https://docs.modelslab.com) --- *This markdown version is optimized for AI agents and LLMs.* **Links:** - [Website](https://modelslab.com) - [API Documentation](https://docs.modelslab.com) - [Blog](https://modelslab.com/blog) --- *Generated by ModelsLab - 2026-05-13*