NeMo Guardrails

What is NeMo Guardrails?

NVIDIA NeMo Guardrails is a free, open-source Python library that lets you attach programmable guardrails to applications built on large language models. The rails sit between your application code and the model and can block inappropriate, off-topic or malicious inputs and outputs. The documentation covers content safety, jailbreak protection, topic control, detection and masking of personal data, and checks on tool calls for agents. Dialog flows and rules are written as YAML configuration plus Colang, and custom checks can be Python actions or LangChain tools. It works with NVIDIA safety models, community models such as LlamaGuard, and third-party services including ActiveFence, Cisco AI Defense, Presidio and Private AI. You can install it with pip and use the Python SDK or the library's API server for development and testing. For production, NVIDIA offers a NeMo Guardrails microservice inside the NeMo microservices platform, deployed on Kubernetes with Helm and supported under the NVIDIA AI Enterprise plan. The library itself is open source and free to use.

Use cases

• Block unsafe LLM inputs and outputs
• Detect jailbreak attempts
• Keep a chatbot on approved topics
• Mask personal data in prompts and replies
• Validate tool calls from AI agents
• Steer conversations along set dialog flows

Who it's for

• AI developers
• ML engineers
• Security teams
• Enterprise AI teams

Why use NeMo Guardrails

• Safety, topic and security checks live in configuration outside your application logic.
• You can mix NVIDIA models, community models and third-party services for each type of check.
• Configurations built locally with the library carry over to the production microservice.
• Agent guidance covers tool inputs, execution rails and monitoring of agent actions.
• As an open-source package, you can inspect and extend it.

NeMo Guardrails features

• Input, output, dialog and topical rails
• Content safety checks using LLM self-checking or safety models
• Jailbreak detection with self-check and heuristic methods
• PII detection and masking through GLiNER, Presidio and others
• Tool call validation and LangGraph integration for agents
• Colang language and YAML configuration
• Custom Python actions and LangChain tool registration
• Python SDK and API server for guardrailed chat completions
• Portable configs between the library and the microservice

NeMo Guardrails integrations

• LangChain
• LangGraph
• LlamaGuard
• ActiveFence
• Cisco AI Defense
• Microsoft Presidio
• Private AI
• AutoAlign
• GuardrailsAI

How to get started with NeMo Guardrails

1. Install the nemoguardrails package from PyPI in your Python environment.
2. Write a config folder with YAML settings and Colang flows that define your rails.
3. Load it with RailsConfig and LLMRails in your code, or run the library's API server.
4. Test the rails in development, then move to the NeMo Guardrails microservice if you need production-grade scaling.

NeMo Guardrails limitations

• The library's API server has no built-in high availability
• The production microservice requires the NeMo microservices platform
• Microservice support is part of the NVIDIA AI Enterprise plan
• The library is self-managed; community support only outside NVIDIA AI Enterprise

NeMo Guardrails pricing

NeMo Guardrails is open source, so you can use and self-host it for free.

Prices can change and may vary by country or exclude tax. Check the official pricing page.

NeMo Guardrails alternatives

The main NeMo Guardrails alternatives listed on Debuly are Guardrails AI, Lakera Guard, Lunary, AI SDK, Agent Governance Toolkit and Cleanlab Trustworthy Language Model. The table compares pricing and platforms; order is A–Z.

1. 

   Agent Governance Toolkit

   Microsoft's open-source toolkit for enforcing policy, identity, sandboxing and audit records on autonomous AI agents before they reach production.

   Developer platformsAutomation and agents
   Open sourceVisit Agent Governance Toolkit (official site)
2. 

   AI SDK

   Free open-source TypeScript toolkit for building AI-powered apps and agents with one API across OpenAI, Anthropic, Google and more.

   Developer platformsCodingAutomation and agents
   Open sourceVisit AI SDK (official site)
3. 

   Cleanlab Trustworthy Language Model

   API that gives any LLM response a trustworthiness score in real time, so you can catch likely errors before users see them.

   Developer platformsAutomation and agents
   Visit Cleanlab Trustworthy Language Model (official site)
4. 

   Guardrails AI

   Open-source framework that wraps LLM apps with validators for PII, hallucination and policy checks, plus Snowglobe simulation testing.

   Developer platformsCodingAutomation and agents
   Open sourceVisit Guardrails AI (official site)
5. 

   Lakera Guard

   Runtime AI guardrails API that screens LLM and agent traffic for prompt attacks, data leakage and unsafe content, with a free Community tier.

   Developer platformsAutomation and agents
   FreemiumVisit Lakera Guard (official site)
6. 

   Lunary

   LLM observability and evaluation platform for logging calls, tracing agents, managing prompts and running evaluations.

   Developer platformsAutomation and agentsCoding
   PaidVisit Lunary (official site)

NeMo Guardrails FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: from docs.nvidia.com · fetched 24 Oct 2026 · Vendor trademark, used only to identify the product; removed on request
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.