Do you need an llms.txt file?

The format was proposed by Jeremy Howard in September 2024. It is a Markdown file at the root of a site — /llms.txt — with the site's name as a heading, a short summary, and lists of links to the pages that matter, each with a description. The idea is to give an AI tool a clean map of a site instead of making it parse navigation and layout.

Its author describes it as mainly for use at inference time: an assistant fetching it on demand while helping someone, rather than a crawler collecting training data.

Google's optimisation guide for AI search features is explicit. You do not need new machine-readable files, AI text files or Markdown to appear in Google Search, and Google Search itself does not use them. Creating one is fine for other services, but it will neither help nor harm your visibility or rankings in Google.

Earlier, Google's search advocates had compared the file to the old keywords meta tag. The guide now puts that in writing.

Chrome's Lighthouse added an experimental Agentic Browsing category that includes an llms.txt check. That is a different product with a different purpose — how well a page works for browser-based agents — and it is separate from Lighthouse's SEO audits. A missing llms.txt in that category says nothing about how Google ranks you.

The file is most useful where AI tools are genuinely reading your content to do work: developer documentation, API references, software products whose users ask assistants how to use them. There, a maintained map of the docs can help a coding assistant find the right page. If it is generated from your existing documentation, it costs little to keep.

For a typical business website — services, about, contact — there is no evidence it changes whether you are found or cited.

Kramiva's own website has no llms.txt — you can check by visiting /llms.txt and getting a 404. The time went into the things above instead.