Generate a valid llms.txt from your sitemap
Real page titles and descriptions, grouped into sections, straight from your sitemap. Nothing invented, no model involved.
What llms.txt is, and isn't
It's a proposed convention: a markdown file at your site root that gives an AI system a curated map of your important pages. It is not robots.txt -- it doesn't control crawl access -- and there's no evidence yet that having one changes whether you rank or get cited. We think it's worth having anyway, since it costs nothing to publish, but we won't overclaim what it does for you.
How it's generated
We discover your sitemap
robots.txt's Sitemap: directive first, then common paths as a fallback. We handle sitemap-index files and gzipped sitemaps, and cap at the 200 shallowest URLs on larger sites.
We pull real titles and descriptions
Each page is fetched concurrently (a small pool, a 60-second budget) for its actual <title> and meta description -- never invented. A page with no description gets left blank, not filled in with guessed copy.
Template-driven, not model-written
Grouped into sections by path (/docs/, /blog/, /features/), rendered straight into valid llms.txt markdown. No LLM call anywhere in this pipeline -- what you get is a strong first draft to edit, not a finished document.
Frequently asked questions
What is llms.txt actually for?
A proposed convention for giving an AI system a curated, markdown-formatted map of a site's important pages -- roughly analogous to a sitemap written for a language model to read, rather than a search engine.
Does having an llms.txt help me rank or get cited more?
There's no evidence of that, and we won't claim otherwise. It's a convenience file, not a ranking signal, and it is not a replacement for robots.txt -- robots.txt controls crawl access; llms.txt is just a curated index for crawlers that choose to read it.
Which AI crawlers actually read llms.txt today?
Adoption is uneven and changing -- some AI products reference it, most crawlers don't fetch it as part of their normal indexing behavior yet. Publishing one costs nothing and may help going forward; it's not a bet you should expect an immediate return from.
Will it write my page descriptions for me?
No. This tool never calls a model and never invents copy -- if a page has no meta description, the generated file simply omits one for that entry rather than guessing.
Why do you need my email?
It doesn't require a signup, a credit card, or any GetBotRank credits -- just a domain and an email. The generated file renders on this page immediately, and we send you a copy plus a shareable link at the same time.