What llms.txt is, and whether your site needs one
A plain-text map of your site for language models. It takes ten minutes to write. Here is what goes in it and what it does not do.
1 min read
llms.txt is a Markdown file at the root of a site, next to robots.txt. Where robots.txt says what a crawler may read, llms.txt says what is worth reading: a short description of the site and links to the pages that explain it best.
What goes in it
The format is deliberately loose. Most good ones follow the same shape:
- A title, and one paragraph on what the site or product is.
- Sections of links, each with a line on what the page answers.
- Optionally, a link to a longer file with the full text of the key pages.
# Example Co
> Scheduling software for dental clinics.
## Docs
- [Quickstart](https://example.com/docs/quickstart): set up a clinic in ten minutes
What it does not do
It does not grant or block access; that is still robots.txt. It is not a ranking signal anyone has promised to use. Think of it as a well-labelled front door, for tools that choose to use it.
Should you add one
If you have documentation, a product people ask assistants about, or a blog with answers worth quoting, yes. It is cheap, it is plain text, and it forces a useful exercise: deciding which ten pages explain your business.
Keep it current. A map that points at pages you deleted is worse than none.
- #llms-txt
- #ai-crawlers
Updated .
