Free tool

llms.txt generator

Build a valid llms.txt for your site: the short, curated map that tells an AI assistant which of your pages matter and when to use them. Fill the fields, copy the file, drop it at your root.

Becomes the H1 — the file's title.

The first thing a model reads. Say what you do, not how great it is.

llms.txt
#
  • The site name is required — it becomes the H1.
  • The summary is required. It is the one line a model reads first.
  • Add at least one link — a file with no links tells a model nothing.

Where to put it

Save the file as llms.txt at the root of your domain, so it resolves at https://yoursite.com/llms.txt. It is a plain static file — no build step, no header needed.

A file is not a strategy

llms.txt helps a model that already reached your site. It does nothing about whether assistants name you when a buyer asks about your category in the first place. That is worth measuring before you optimise anything.

Run the free audit

What llms.txt is, and what it is not

llms.txt is a markdown file at the root of a domain that lists the pages an AI assistant should read to understand a site. It was proposed by Jeremy Howard in September 2024, and the format is deliberately small: an H1 with the site name, a blockquote summary, then H2 sections holding markdown links, each with an optional note after a colon.

The problem it addresses is context, not access. A model answering a question about your product cannot read your whole site; it reads what it can find and fit. llms.txt is your chance to say which pages carry the answer.

llms.txt vs robots.txt

They are often confused and they solve different problems. robots.txt is a permission file — it tells a crawler what it may fetch, and blocking GPTBot there means ChatGPT cannot read your pages at all. llms.txt assumes access already exists and directs attention within it.

The order matters: a perfect llms.txt on a site that blocks AI crawlers achieves nothing, because no model ever fetches the file. If you have not checked, our AI crawler access checker reads your live robots.txt and tells you in a few seconds.

An honest note on adoption

No major AI provider has publicly committed to reading llms.txt the way search engines committed to robots.txt. Adoption is real but concentrated among developer-tool and documentation sites. We publish one, and we would not claim it is why anyone gets cited. The case for it is that it costs a few minutes and cannot hurt — not that it is a ranking factor.

What a good file looks like

Three habits separate a useful file from a wasted one.

  • Write the notes. The text after the colon is what tells a model when a page is relevant. A list of bare links is a sitemap with extra steps.
  • Keep it short. Ten to forty links beats two hundred. The whole point is fitting in a context window alongside the question being asked.
  • Use absolute URLs. A model fetches these directly and has no page to resolve a relative path against.

Frequently asked questions

What is llms.txt?
llms.txt is a plain markdown file at the root of a domain that lists the pages an AI assistant should read to understand the site. It was proposed by Jeremy Howard in 2024. Where robots.txt says what a crawler may fetch, llms.txt says what is worth reading — a curated map sized to fit a model context window.
Is llms.txt the same as robots.txt?
No. robots.txt is a permission file: it grants or denies crawler access, and crawlers respect it by convention. llms.txt is a discovery file: it points to your most useful pages and describes them. One controls access, the other controls emphasis, and a site can want both.
Do AI models actually read llms.txt?
Support is uneven and worth being honest about. No major provider has committed to reading it the way search engines commit to robots.txt. Adoption is growing among developer-tool sites, and several documentation platforms now generate one by default. The file costs almost nothing to publish, which is the main argument for it today.
What is the difference between llms.txt and llms-full.txt?
llms.txt is the index: a short list of links with one-line descriptions. llms-full.txt is the content itself — the full markdown of those pages concatenated into one file, so a model can read everything without following links. Publish llms.txt first; llms-full.txt only makes sense once the index is stable.
How many links should llms.txt contain?
Fewer than you would put in a sitemap. The point is fitting in a context window, so most sites are best served by 10 to 40 links covering documentation, key product pages and the handful of articles that define their expertise. A file listing every URL defeats its own purpose.
Where does the file go?
At the root of the domain, so it resolves at https://yoursite.com/llms.txt. It is a static text file — no build step, no special headers, no server configuration.