What is llms.txt?
llms.txt is a plain-text file at a site's root that lists its most important pages for AI systems, written in Markdown.
Published
Definition
llms.txt is a plain-text file placed at the root of a website that lists its most important pages for AI systems, written in Markdown. It was proposed in 2024 as a convention, not a standard.
Where robots.txt defines what a crawler is allowed to read, llms.txt suggests what is worth reading.
Why It Matters for AI Visibility
A crawler landing on a large site has no map. It sees navigation, footers, pagination and tag archives, and has to work out which pages carry the substance. llms.txt hands it the answer directly: here are the twenty pages that matter, each with one line explaining what it is.
It is a hint, not an instruction — nothing obliges a model to read it. But it costs almost nothing to publish, it doubles as an honest inventory of your own site, and the one-line descriptions are themselves clean, quotable summaries.
Example
# example.com — Developer docs
> REST API documentation for the example.com platform.
## Getting started
- [Quickstart](https://example.com/start): Authenticate and make a first call.
- [Authentication](https://example.com/auth): API keys, scopes, rotation.One caveat: the file is not inherited by subdomains. Each subdomain needs its own at its own root.
Related
- robots.txt · Content-Signal · Agent readiness
- Tool: llms.txt Generator
- Article: llms.txt vs robots.txt