What is an llms.txt file?

It is a Markdown-formatted text file placed at yoursite.com/llms.txt that lists your most important pages and gives AI engines a clean, curated map of what to read. It is a proposal to crawlers, not a rule they must obey.

Does llms.txt guarantee an AI engine will cite me?

No. No file can promise a citation. An llms.txt can make your important pages easier to find and read, but whether an engine surfaces you depends on the page content and the engine's own choices. You can measure readiness, not guarantee the outcome.

Is llms.txt the same as robots.txt?

No. robots.txt tells crawlers what they may not access. llms.txt does the opposite: it highlights what you want read, in a format aimed at language models rather than search bots.

← Back to glossary

Glossary · 2 min read

llms.txt

A plain-text file at the root of your site that points AI engines and crawlers to the pages and content you most want them to read.

You probably already asked an AI engine whether you need an llms.txt file. It told you to add one and move on.

Here is what that answer left out: a file pointing at your pages does nothing if the pages themselves cannot be read.

What llms.txt actually is

An llms.txt is a single plain-text file at the root of your site, at yoursite.com/llms.txt. It lists the pages you most want AI engines to read, in a clean Markdown format, so a model is not left guessing which of your hundred URLs matter.

Think of it as a curated table of contents written for machines. Your navigation is built for a human clicking around. An llms.txt is built for a crawler that wants the short list.

What it does, and what it does not

It can help. A model fetching your site gets a direct pointer to your real answers instead of digging through layout, menus, and archive pages.

It is not a command. An llms.txt is a proposal. Engines can read it, ignore it, or treat it as one signal among many. That is the honest limit of the format.

And it is not a substitute for readable pages. If the pages your llms.txt points to are a wall of layout with the answer buried on line 70, the file just guides a machine straight to something it still cannot parse.

The damaging admission

Most llms.txt files are published once and never checked again. The site adds forty new pages, changes its pricing, retires a product, and the file still points at the old map.

A pointer to a stale or unreadable page is worse than no pointer at all. The value is not in having the file. It is in the file being accurate, and in the pages it names being legible to a machine. Both drift the moment you keep publishing.

How to check yours

Pick the one page you would most want an engine to quote. Confirm it is listed in your llms.txt, then check whether a machine can actually read that page: a clear heading, a direct answer in plain text, and schema that labels what the page is.

That turns "I added an llms.txt" into "the three pages it points to actually read cleanly, and here is the one that does not." You can run a free scan on any URL to see how an engine reads it today, or read how Citedon makes a site agent-ready and keeps it that way as your content changes.

See whether engines can actually read your key pages.

Run a free scan. No signup. You get a readiness score and the gaps to fix, in about a minute.

Run a free scan How it works