Free tool
llms.txt generator
An index an assistant can read: what your site is, and where the pages that answer questions actually live. Built from a live crawl, so every link in it is a page that answered.
Want this running 24/7 on your account?
Free tool
An index an assistant can read: what your site is, and where the pages that answer questions actually live. Built from a live crawl, so every link in it is a page that answered.
Want this running 24/7 on your account?
It reads your robots.txt, finds your sitemap if you publish one, and fetches up to ten pages. For each page it keeps the URL, the title and the meta description, then sorts them into product, pricing, documentation, writing, company and legal by their path.
The output is one markdown file. Pages that returned anything other than a 200 are excluded and listed underneath with the reason, so a gap in the index is a decision you can see rather than a silent omission.
Save it as llms.txt at the root of your domain, next to robots.txt, so it resolves at yoursite.com/llms.txt. Most frameworks serve anything in the public or static directory from the root; on a CMS it is usually a redirect or a file upload rather than a page.
It is worth publishing only if an assistant can reach your pages in the first place, so run the AI crawler access check as well. A perfect index behind a blanket robots.txt block helps nobody.
Whenever the shape of the site changes: a new section, a renamed pricing page, a documentation move. The file is an index, so the failure mode is staleness rather than wrongness, and a stale index is worse than none once it starts pointing at pages that have moved.
Last updated August 31, 2026. All 17 calculators and AI helpers are listed on the free tools hub.
FAQ
A plain markdown file at the root of a site that tells an assistant what the site is and links the pages worth reading, grouped by purpose. It is to AI assistants roughly what a sitemap is to a search crawler: not a ranking signal, but a shortcut past guessing from your navigation.
Because a hand-written index goes stale and nobody notices. The most common defect is a link to a section root that was never a page, which sends an assistant to a 404 and teaches it that the site is broken. This generator can only link a URL it just fetched and got a 200 from, so that defect is structurally impossible.
Up to ten, four at a time, starting from your sitemap where you have one and from your home page where you do not. Your robots.txt is obeyed. If your site has more than ten pages the file covers the ones closest to the home page and says so.
No. The summary line is your home page's own meta description, or its title when there is no description, and it is left out entirely when there is nothing to quote. Each link carries that page's own meta description or nothing at all. No sentence in the output was written by a model.