What is llms.txt, and should you add one?

You have seen llms.txt mentioned and cannot tell whether it is a standard or a fashion.

What it is

A plain-text file at your root describing your site in language a model can read: what it is, what the main pages are, what is true about it. Proposed by convention, not by any standards body, and adoption is limited.

It is not a directive

Unlike robots.txt, nothing is obliged to read or respect it. Anyone telling you it controls how models use your content is wrong. It is a courtesy map.

Add it anyway, for the reason that it costs an hour

Writing a plain description of what your site is and what is true about it is worth doing regardless, and having it in one machine-readable place cannot hurt. Just do not build a strategy on it.

What it is and who reads it

A proposed convention: a markdown file at your root summarising your site for language models. It is a proposal, not a standard, and adoption by the major model providers is limited. Anyone telling you it is required is describing a future that has not arrived.

Cost, benefit, and the honest verdict

It takes twenty minutes and cannot hurt you, so add it if you have twenty minutes free. Do not reorganise anything around it, do not pay for it, and do not expect measurable traffic. Judged as a lottery ticket with a near-zero price it is fine. Judged as a strategy it is a distraction from the fact that nothing links to your site.

Measured, not asserted

This project publishes one, generated at build time from the same route data as its sitemap, so it cannot drift from the site it describes.

Free tool for this: Schema Inventory. No account, nothing uploaded.

Where this goes deeper

Every number on this page comes from one complete dataset: one product taken from zero to 100K+ users on search alone, with nothing spent on advertising. The full argument is Chapter 12 of the book. Five chapters are free to read.