Free tool

Generate llms.txt

Enter your domain. We read robots.txt and your sitemap, fetch up to 60 pages, and write the file from your titles and descriptions. Result in under two minutes, with a grade of the file we produced.

We read your robots.txt and sitemap, scan up to 60 pages, and build the file from your existing titles, descriptions, and headings. No AI-written copy. No account required.

1Discover

robots.txt first (we skip what it disallows for us), then sitemap.xml or the nav links on your home page. Localised sites keep one language.

2Rank and fetch

Home, docs, API, product, pricing, blog (recent ten), about. Family caps stop one large blog crowding out documentation. 2 requests per second, 8 s timeout, 2 MB per page.

3Emit

H1 site name, blockquote summary from your home page, H2 sections with - [title](url): description lines, an Optional section for legal and low-priority pages. Same site in, same file out.

What to edit afterwards. The blockquote is your meta description; rewrite it as a paragraph for a reader who has never heard of you. Descriptions truncated at 160 characters end with an ellipsis; finish the sentence. Pages with no description are listed without one, which the validator accepts.

llms-full.txt. Tick the box to also get a companion file with each page's text inlined (capped at 500 KB). It is not part of the llmstxt.org proposal, just a convention some sites use; the validator checks only llms.txt.

Already have a file? Validate it instead.