llms.txt: what it is and how to build one
An llms.txt is a Markdown file at your-domain.com/llms.txt that summarises a website for language models and links its most important pages. It is a proposal from llmstxt.org, not a standard: no AI service is obliged to read it. It is quick to create and complements robots.txt and sitemap, but replaces neither.
Structure per llmstxt.org
# Name
The only required part: an H1 with the name of the site or project.
> Summary
A blockquote with the key information needed to understand the rest of the file.
## Sections
H2 sections with link lists: “- [Title](URL): note”. A section named “Optional” marks links that can be skipped.
Example
# Example Ltd > Web design and SEO for small businesses. ## Pages - [Home](https://www.example.com/): What we do and for whom - [Contact](https://www.example.com/contact): Address, phone, booking ## Services - [Web design](https://www.example.com/services/web-design): Websites at a fixed price - [SEO](https://www.example.com/services/seo): Search engine optimisation
llms.txt: frequently asked questions
What is an llms.txt?
A Markdown file at the root of a website (/llms.txt) that gives language models a short overview: the name as a heading, a summary as a quote, then sections with links to the most important pages. The format was proposed in 2024 on llmstxt.org.
Is llms.txt a standard? Do ChatGPT and Google read it?
It is a proposal, not a standard. Whether and how an AI service evaluates the file is up to each vendor; no one is obliged to. llms.txt does not replace robots.txt (who may crawl) or the sitemap (which pages exist).
How does the generator choose the pages?
It reads the Sitemap lines in your robots.txt, otherwise /sitemap.xml; for a sitemap index it follows one more level and prefers page sitemaps over posts. Of the addresses on your own domain it takes the start page and the pages with the shortest paths, at most 25, and leaves out files, tag and archive pages and pages set to noindex. Without a sitemap it uses the links on your start page.
Where do I put the file?
As a plain text file named llms.txt in the root directory of your website, so that it is reachable at your-domain.com/llms.txt. On WordPress that is the folder containing wp-config.php; with other systems ask your host or agency.
Why are some pages missing or poorly described?
The generator reads the HTML your server delivers. Content that only appears through JavaScript is not visible to it, and a page without a meta description gets its first longer paragraph instead. Please read the result and edit it before uploading.