llms.txt is a proposed convention: a plain-markdown file at yoursite.com/llms.txt that says what the site is and links to the pages worth reading. It is written for a language model rather than a browser — no navigation, no marketing, just the summary and the links.
It is not a standard, and no assistant is obliged to read it.
A heading with the site's name, a blockquote summarising it in a sentence or two, then linked sections. Ours lists what the product is, the pages that explain it, and how to get in touch.
# Your Business
> One or two sentences saying what you do and who for.
## Pages
- [What we do](https://yoursite.com/services): the services, plainly listed.
- [Contact](https://yoursite.com/contact): how to reach us.It is cheap and it does no harm, so if you have ten minutes, write one. But be clear about what it is worth relative to everything else:
We check for it, and we report it as present or absent rather than scoring you on it.
The failure mode is drift: a file written once, linking to pages that have since moved. A link in llms.txt that leads nowhere is worse than no llms.txt, because the one reader it has is a machine that will not look for the new URL.
No. It is a proposed convention. Nothing requires an assistant to fetch or honour it.
No. robots.txt controls access; llms.txt describes content. They answer different questions and both belong at the site root.
No. Schema.org markup is read by search engines and assistants alike and attaches facts to specific pages; llms.txt is a site-level summary.
At the root of the domain — https://yoursite.com/llms.txt — served as plain text or markdown.
Where does your site stand? The scan is free, takes about 60 seconds, and needs no account. It fetches your homepage exactly the way GPTBot, ClaudeBot and PerplexityBot do and tells you what they receive.