NeMo Guardrails review

Q: How do I install NeMo Guardrails?

Run pip install nemoguardrails. The nemoguardrails package needs Python 3.10 or newer, below 3.14.

NeMo Guardrails is an open-source Python toolkit for adding programmable guardrails to LLM-based conversational applications.

Updated Nov 1, 2026. Review written Oct 22, 2026 from the project's own sources; usage data refreshed every Monday.

TL;DR

NeMo Guardrails is an open-source Python library from NVIDIA that places programmable guardrails between application code and an LLM. It is for developers building chatbots, RAG systems and LLM endpoints who need to block, alter or validate inputs and outputs. Rails are configured in YAML and Colang, a dialogue modeling language introduced by the toolkit.

7.4kGitHub stars
+47new stars, week of Oct 25
109.3kdownloads a week, −18% vs 4-week avg
#18in Evals and observability this week

NeMo Guardrails at a glance

Category: Evals and observability
Maintained by: NVIDIA-NeMo (organization, @NVIDIA-NeMo on GitHub)
Language: Python, JavaScript, Shell
Repository: github.com/NVIDIA-NeMo/Guardrails
Website: docs.nvidia.com/nemo/guardrails/latest/index.html
PyPI package: nemoguardrails 0.24.1
Latest release: v0.24.1 on GitHub, Sep 16, 2026
First published: Apr 18, 2023

What NeMo Guardrails does

NeMo Guardrails supports five rail types: input, dialog, retrieval, execution and output rails, each applied at a different stage of an LLM interaction.
NeMo Guardrails provides mechanisms against common LLM vulnerabilities such as jailbreaks and prompt injections.
NeMo Guardrails includes built-in guardrails for self-checking, fact-checking, hallucination detection, NVIDIA safety models, and third-party API integrations.
NeMo Guardrails can steer an LLM to follow predefined conversational paths, such as authentication or support procedures.
NeMo Guardrails can mask sensitive data such as person names and email addresses on user input.
NeMo Guardrails ships an evaluation tool, nemoguardrails evaluate, covering topical rails, fact-checking, moderation and hallucination.
NeMo Guardrails runs an OpenAI-compatible server, including a /v1/checks endpoint for running input or output rails without generating a model response.

How it works

NeMo Guardrails loads a configuration folder with config.yml, Colang .co files and optional Python actions. You call it through the Python API with LLMRails and generate/generate_async, or through a guardrails server exposing an HTTP API at /v1/chat/completions. It works with multiple LLMs, including OpenAI models, and optionally wraps LangChain chains. Configurations are portable to the NVIDIA NeMo Guardrails microservice for Kubernetes deployment.

How to install and start using NeMo Guardrails

With pip

pip install nemoguardrails

Requirements

Python 3.10, 3.11, 3.12 or 3.13.
Access to a supported LLM, such as an OpenAI model configured in config.yml.

First steps

Create a config folder with a config.yml defining the LLM model and active input and output rails.
Add Colang .co files to define dialog rails, such as greeting flows or responses to insults.
Load the configuration with RailsConfig.from_path and create an LLMRails instance, then call generate with your messages.
Alternatively, start a server with nemoguardrails server [--config PATH/TO/CONFIGS] [--port PORT].

Who NeMo Guardrails is for

Teams building RAG question-answering systems that need fact-checking and output moderation.
Developers of domain-specific chatbots that must stay on topic and follow designed conversational flows.
Teams adding safety guardrails around custom LLM endpoints or LangChain chains.

Limitations to know

NeMo Guardrails' built-in guardrails may or may not suit a given production use case, according to the README.
NeMo Guardrails requires learning Colang, its own dialogue modeling language, to define dialog rails.
NeMo Guardrails collects anonymous usage stats unless you opt out, for example with NEMO_GUARDRAILS_NO_USAGE_STATS=1 or DO_NOT_TRACK=1.
NeMo Guardrails' LangChain integration is opt-in and requires installing the LangChain packages your configuration needs.

Is NeMo Guardrails growing? Usage data

In the week of Oct 25–31, 2026, NeMo Guardrails ranked #359 on the Devtoolwatch Board and #18 in Evals and observability, with a momentum score of 19 out of 100. Its main package had 109.3k downloads that week, −18% against its own 4-week average.

New GitHub stars per week for the last 12 weeks, from GitHub's stargazer history (today's stargazers by the week they starred).

Show the weekly numbers

New GitHub stars per week for NeMo Guardrails
Week of	New stars
Aug 9, 2026	63
Aug 16, 2026	41
Aug 23, 2026	37
Aug 30, 2026	43
Sep 6, 2026	38
Sep 13, 2026	60
Sep 20, 2026	35
Sep 27, 2026	44
Oct 4, 2026	36
Oct 11, 2026	51
Oct 18, 2026	42
Oct 25, 2026	47

On Hacker News

CUDA 11.0, 182 points and 65 comments (Jul 8, 2020)
CUDA Toolkit Release Notes, 160 points and 119 comments (Nov 24, 2019)
Tile IR Specification, 19 points and 3 comments (Dec 4, 2025)
Nvidia uses raytracing to model 5G network design, 6 points and 0 comments (Jul 4, 2024)
CUDA 13 release notes [pdf], 5 points and 0 comments (Sep 5, 2025)

The Board ranks tools by four signals each week: new stars, new stars for the repo's size, added downloads and download growth beyond the market. How the ranking works.

NeMo Guardrails alternatives

Other LLM evals and observability tools on the Board in the week of Oct 25–31, 2026, by momentum:

Evals and observability ranked by momentum, the week of Oct 25–31, 2026
Rank	Tool	Stars	Downloads/wk
#1	SkillSpectorSkillSpector is NVIDIA's security scanner that checks AI agent skills for vulnerabilities, malicious patterns, and supply-chain risks…	23.7k	–
#2	JudgevalJudgeval is an open-source Python SDK for tracing and evaluating LLM-powered agents with prompt-based judges and online monitoring.	1.1k	80.2k
#3	A.I.G (AI-Infra-Guard)A full-stack AI Red Teaming platform securing AI ecosystems via Agent Scan, Skills Scan, MCP scan, AI Infra scan and LLM jailbreak…	7.2k	–
#4	LangfuseLangfuse is an open source LLM engineering platform for tracing, evaluating, prompt management, and debugging LLM applications and agents.	36.8k	9.8M
#5	OpikOpik is an open-source platform from Comet for tracing, evaluating, and monitoring LLM applications, RAG systems, and AI agents.	23.1k	709.1k
#6	promptfoopromptfoo is a CLI and library for evaluating and red-teaming LLM-based apps, including prompts, agents, and RAG pipelines.	26.7k	1.2M

All LLM evals and observability tools, ranked

Recent NeMo Guardrails releases

v0.24.1, Sep 16, 2026
v0.24.0, Aug 26, 2026
v0.23.0, Jul 1, 2026
v0.22.0, May 22, 2026
v0.21.0, Mar 12, 2026

NeMo Guardrails FAQ

Is NeMo Guardrails free and open source?

NeMo Guardrails's code is on GitHub (NVIDIA-NeMo/Guardrails), but the repository doesn't declare a standard open-source license. Check its license file before using it commercially.

How do I install NeMo Guardrails?

Run pip install nemoguardrails. The nemoguardrails package needs Python 3.10 or newer, below 3.14.

Is NeMo Guardrails still maintained?

Yes. The NVIDIA-NeMo/Guardrails repository was last updated on Oct 21, 2026, and the latest release, v0.24.1 on GitHub, came out on Sep 16, 2026.

How popular is NeMo Guardrails?

NeMo Guardrails has 7.4k GitHub stars as of Nov 1, 2026. It gained 47 new stars in the week of Oct 25–31, 2026, against a 12-week average of 45 a week. Its main package had 109.3k downloads that week, −18% against its own 4-week average. That made it #18 among LLM evals and observability tools on the Devtoolwatch Board.

What are the best alternatives to NeMo Guardrails?

Other LLM evals and observability tools on the Devtoolwatch Board in the week of Oct 25–31, 2026, by momentum: SkillSpector (#1), Judgeval (#2), A.I.G (AI-Infra-Guard) (#3), Langfuse (#4), Opik (#5). The alternatives table above has their numbers.

What is NeMo Guardrails?

NeMo Guardrails is an open-source NVIDIA toolkit for adding programmable guardrails to LLM-based conversational applications. Guardrails control model behavior, such as avoiding certain topics, following dialog paths, or extracting structured data.

How does NeMo Guardrails work?

NeMo Guardrails sits between your application and the LLM. A YAML and Colang configuration defines the model and rails that check or alter user input, dialog, retrieved chunks, tool calls and model output. You call it through the Python API or an HTTP server.

Which LLMs does NeMo Guardrails support?

NeMo Guardrails works with multiple LLMs, including OpenAI GPT-3.5, GPT-4, LLaMa-2, Falcon, Vicuna and Mosaic. Version 0.22.0 made LangChain optional and added built-in support for OpenAI-compatible LLMs.

What is Colang?

Colang is a modeling language introduced by NeMo Guardrails for designing dialogue flows. It has Python-like syntax. Versions 1.0 and 2.0 are supported, and 1.0 is the default.

Can NeMo Guardrails run as a server?

NeMo Guardrails includes a CLI that starts a guardrails server exposing an HTTP API. The server is OpenAI-compatible and can also run in a Docker container built from the provided Dockerfile.

Sources

NeMo Guardrails README badge

Maintainers can show the tool's weekly Board rank in their README:

[![Devtoolwatch](https://index.devtoolwatch.com/badge/nemo-guardrails.svg)](https://index.devtoolwatch.com/tools/nemo-guardrails/)

How this page is made: Devtoolwatch writes each review from the project's README, releases, package metadata and website, with an AI model, then checks every claim against the source it cites and drops what the sources don't support. The numbers come from Devtoolwatch's daily GitHub snapshots, GitHub's stargazer history, and the npm and PyPI download APIs. Spotted an error? Email [email protected].