All posts

How to Check if a Site Has an llms.txt File (And If It Follows the Format)

··3 min read

To check if a site has an llms.txt file, open /llms.txt at its root and see whether it is a Markdown file that starts with an H1. The llms.txt checker fetches it for up to 100 sites a run (500 after a free Google sign-in), checks it against the llmstxt.org proposal and reports the title, summary, sections, link counts and whether a companion /llms-full.txt exists.

It makes plain requests for two public files per site and nothing else.

The by-hand way

  1. Visit https://yoursite.com/llms.txt. A real file is plain text or Markdown.
  2. Check the first line is # Site or project name. That H1 is the only required part.
  3. Look for a > summary line, then H2 sections with link lists such as - [Docs](https://example.com/docs.md): what it covers.
  4. Try /llms-full.txt too.

Beware the false positive: many sites answer an unknown path with their home page or an HTML 404 and a 200 status. It looks like a file but is not one.

Step 1: paste the sites

Paste one website or a list, one per line. Up to 100 a run without an account, 500 after a free Google sign-in.

Step 2: run the check

Each site's /llms.txt is fetched, HTTPS first, along with /llms-full.txt on the same origin. The first 1 MiB of the file is read.

Step 3: read each verdict

Verdict What it means What to do
Follows the format H1 title and the expected structure are present Nothing to fix
Partly follows it Something is there but parts of the structure are missing Read the notes and fill the gaps
Not llms.txt format The response is not a usable Markdown file, often an HTML page Publish a real file
No llms.txt Nothing at the path Optional: add one
Unreachable The site did not answer Try again later

Does llms.txt matter

It costs little to publish and is mainly useful to tools that fetch documentation for coding assistants. Two things matter more for appearing in AI answers: that the crawlers can fetch your pages at all, which robots.txt controls (see checking robots.txt for AI crawlers), and that the pages linked from llms.txt are real and current. You can check those rules across a list with the robots.txt checker.

What this check cannot do

  • llms.txt is a proposal. The checker cannot say whether any given assistant reads the file.
  • Linked pages are not fetched. Only the link form is checked.
  • The first 1 MiB of the file is read.
  • The limit is 100 sites a run without an account and 500 signed in.

For other site files, see checking a sitemap for errors.

Check the people behind the sites

If you are emailing the owners of these sites, check the list first. Upload it on the verify page: the first 100 email checks are free with a Google sign-in, no card. Each address comes back valid, risky or invalid with a reason, and nothing is sent to anyone on the list. For larger or recurring lists, see bulk email verification, list cleaning and pricing.

Check your sites now

Paste them into the llms.txt checker.

Common questions

How do I check if a site has an llms.txt?

Open /llms.txt at the root of the site. It should be a Markdown file that starts with an H1 naming the site. An llms.txt checker fetches it, reads its title, summary, sections and links, and says whether it follows the format.

What does a valid llms.txt contain?

An H1 with the site's name, which is the only required part, then optionally a blockquote summary, plain paragraphs, and H2 sections that each list links in the form - [Page name](url): what it covers.

Do ChatGPT, Claude or Google read llms.txt?

It is a proposal, not an IETF or W3C standard, and reading it is up to each operator. The checker reports the file's shape, not whether any assistant uses it.

What is llms-full.txt?

A common convention: the full text of the linked pages in one file. It is not part of the core format, so the checker reports whether it exists without counting it against the site.

Why does a site show Not llms.txt format?

Often the site answers /llms.txt with its HTML 404 or home page and a 200 status, which looks like a file but holds nothing to read.

Check your list before you send

Real mailbox checks over SMTP, and nothing is sent to the address. Sign in with Google and your first 100 email checks are free, no card. Plans for bigger lists are on the pricing page.