19 July 2026

llms.txt: What It Is and Whether Your Site Needs One

The short answer

llms.txt is a proposed convention (introduced by Jeremy Howard in 2024) for a markdown file at your site root, /llms.txt, that gives AI systems a concise, curated summary of your site: what it is, plus annotated links to your most important pages. Unlike robots.txt, which controls crawler access, llms.txt is informational. It is not an official standard, and major AI providers have not confirmed using it, so treat it as a low-cost, low-certainty addition, not an AEO requirement.

Websites are built for browsers: navigation chrome, scripts, and layout that an AI reading your site has to wade through. The llms.txt proposal offers a shortcut: one plain markdown file at the root of your domain that tells language models what your site is and where the substance lives.

What goes in an llms.txt file?

An H1 with the site or project name.
A short blockquote summary of what the site offers.
Sections of markdown links to key pages, each with a one-line description.
Optionally an llms-full.txt containing expanded content in one file.
Plain markdown throughout; it is meant to be read by models, not rendered.

Is llms.txt an official standard?

No. It is a community convention, and that honesty matters for planning. robots.txt is respected because crawlers are built to check it; llms.txt has no equivalent guarantee, and major AI companies have not publicly committed to reading it. Some documentation platforms and developer tools have adopted it, which keeps the convention alive, but claims that it is required for AI visibility are ahead of the evidence.

If your site has real reference value (products, documentation, guides), the cost is twenty minutes and the downside is zero, so it is a reasonable bet; just do not let it substitute for the things that demonstrably matter: clean structure, answer-first content, and structured data. Those fundamentals are covered as part of the launch and SEO module in the Premium Website Masterclass. It is also worth noting that simple sites are inherently easy for AI to read: a single-file template like the Cobalt SaaS Template ships semantic HTML with no script-rendered content, which does more for machine readability than any companion file.