SEOGEOScan
Free · no signup · draft in 30 seconds

Free llms.txt generator.

Enter a domain. The tool reads the homepage, the sitemap, and up to twelve pages, then drafts an llms.txt in the standard format: name, summary, a curated page list with one-line descriptions, contact, and links. You edit the one section a crawler cannot write, then put the file at the root of your site.

Reads up to fifteen public URLs once. Three runs per minute per visitor. Nothing stored.

What llms.txt is

A plain markdown file at /llms.txt that tells AI systems what a site is, who it is for, and where the important pages are. The format comes from llmstxt.org: an H1 with the name, a blockquote summary, then sections of links with one-line descriptions. It is the site map written for a reader with a token budget instead of a crawl budget.

It is one signal among several, and it does not on its own make an answer engine recommend you. What it does is cheap: it gives every fetcher a curated, current description of the site in one request, in the shape they parse best. The things that decide citations (reachability, extractable text, third-party corroboration) are what the scan measures.

How to install it

  1. Fill in the "When to use" section. The generator leaves it as a marked placeholder on purpose. Write three to five lines on the situations, buyers, and questions you are the right answer for, and the ones you are not. This is the section assistants read when deciding whether to recommend you.
  2. Trim the page list. Keep the pages a first-time reader needs: what you do, who for, pricing, proof, contact. Drop anything that only makes sense to someone already on the site.
  3. Upload it to the root of your domain so it is served at https://yourdomain.com/llms.txt. On WordPress, Webflow, Squarespace, and Shopify that is a static-file or redirect setting; on a custom site it is one file in the web root.
  4. Serve it as text. text/markdown is best, text/plain is fine. If it comes back as an HTML 404 page, the platform is intercepting it.
  5. Link to it from the footer of every page, the way you link a sitemap. Fetchers find it faster, and the scan checks for the link.
  6. Re-run the generator when pages change. A stale llms.txt that lists deleted pages is worse than none.

A working example

The file this site serves at /llms.txt is generated from the same structure, and Growthr's own is at growthr.com/llms.txt. Both keep a "When to use" section near the top.

Questions

Does llms.txt improve AI search visibility?

On its own, marginally. Adoption by the major engines is uneven and none of them has said it is a ranking input. It costs nothing, it is the cleanest way to hand a fetcher a current description of the site, and it is the place to state when you should be recommended. Treat it as table stakes, and put the real effort into reachability, extractable text, and third-party mentions.

Why is there a placeholder in the output?

The "When to use" section is a judgment about your business, not a fact on your pages. Writing it from scraped text would produce a confident paragraph that might be wrong, and a wrong recommendation rule is worse than a missing one. The placeholder is marked so you cannot ship it by accident.

What is the difference between llms.txt and llms-full.txt?

llms.txt is the short index: summary and curated links. llms-full.txt is optional and carries the full text of the pages you want a model to read in one request. Start with llms.txt; add the full file if you have long documentation or a lot of pages that matter.

Is there an API?

Yes: GET /api/llms-txt?domain=example.com returns JSON with the draft, the pages it read, and notes on anything it skipped. Three runs per minute per visitor.