llms.txt is a proposed standard (llmstxt.org) for a plain markdown file at the root of a website, at /llms.txt. It gives large language models and AI agents a short, curated map of the site: what it is, and which pages matter most. Think of it as a sitemap written for readers that have limited context windows.
The format
- An H1 with the name of the site or product (the only required part)
- A blockquote with a one or two sentence summary
- Optional paragraphs with key facts
- H2 sections containing markdown lists of links, each with an optional note after a colon
- An optional section named "Optional" for links that can be skipped when context is short
Example
# Acme Analytics > Privacy-friendly product analytics for SaaS teams. Free up to 10k events/month. ## Docs - [Quickstart](https://acme.com/docs/quickstart): install in 5 minutes - [API reference](https://acme.com/docs/api) ## Product - [Pricing](https://acme.com/pricing) - [Acme vs Mixpanel](https://acme.com/compare/mixpanel) ## Optional - [Changelog](https://acme.com/changelog)
llms-full.txt
Some sites also publish /llms-full.txt, which contains the actual text of the key pages in one file, so an agent can read everything in a single request.
Does it help?
llms.txt doesn't replace good HTML, crawler access or structured data. It's a low-effort addition that makes your site easier for agents and coding assistants to understand, and it signals that the site is maintained with AI readers in mind.
Generate yours
RankRoot's free scan reads your homepage and drafts an llms.txt you can edit. RankRoot Pro crawls up to 200 pages and generates both llms.txt and llms-full.txt for the whole site.