As people increasingly reach websites through AI assistants instead of search results, a new convention has appeared: llms.txt. It's a plain-text file that gives large language models a clean, curated map of your site. It's young and not yet an official standard — but it's trivial to add and starting to see adoption.
What it is
Served at /llms.txt, the file offers AI tools a concise, human-readable summary of what your site is and links to your most important pages — without them having to parse your full HTML, navigation, and scripts. Think of it as a hand-written table of contents aimed at machines that read Markdown.
The format
It's just Markdown. An H1 with the site name, a short blockquote summary, then sections of links with brief descriptions:
# Example Co
> What the site is, in one sentence.
## Pages
- [Home](https://example.com/): Overview.
- [Docs](https://example.com/docs/): Product documentation.
- [Blog](https://example.com/blog/): Articles and guides.
## Contact
- [Email](mailto:hi@example.com)Some sites also publish an llms-full.txt with the full text content inlined, for tools that want everything in one fetch.
How it differs from robots.txt
robots.txt tells crawlers what they may not access; llms.txt tells AI tools what's worth reading and how it's organised. One restricts, the other guides. They're complementary — you can and probably should have both.
Is it worth it?
Honestly: it costs ten minutes and can't hurt. No major AI provider has committed to reading it universally yet, so treat it as low-effort future-proofing rather than a guaranteed win. If your site has clear "important pages" — docs, products, key articles — write one. If you're not sure whether your site even serves one, SEO Inspector checks for llms.txt (alongside robots.txt and sitemap.xml) on any site you visit.
