llms.txt

What it is

llms.txt is a proposal by Jeremy Howard, first published in September 2024 and revised as v2 in August 2026. It suggests a Markdown file with a title, short description and curated links, so AI agents can find concise, model-friendly content without parsing whole web pages.

How it works

1. Write a Markdown file with your site name, a one-paragraph summary and sections of key links.
2. Publish it at /llms.txt. Version 2 also allows a file at any path, such as /docs/llms.txt, covering the pages under it.
3. Optionally offer clean Markdown versions of key pages at the same URL with .md added, and point to them with standard link relations.

Benefits

• Cheap to create and maintain. The proposal lists documentation platforms that generate one automatically.
• Gives coding assistants and agents a tidy map of documentation, which the proposal says is its heaviest use today.
• Forces you to decide which pages matter most.

Limits and risks

• Not a confirmed signal. It isn’t a confirmed ranking or citation signal for any major answer engine. The proposal notes that AI labs publish llms.txt files for their own developer docs. That isn’t the same as their products reading yours, and we found no vendor documentation saying they do (checked 2026-10-17).
• Not needed for Google. Google says no new machine-readable files or AI text files are needed for AI Overviews or AI Mode.
• Upkeep. It’s another file to keep in sync with the site.
• Agentic search. Christopher Penn’s view is that pages that are easy for agents to read help in agentic search. That’s a practitioner view, not a measured result.