llms.txt is a file placed at the root of a domain (for example, /llms.txt) that offers large language models a clean, curated index of a site: a short description of the organization followed by links to the pages that best represent it. It is analogous to robots.txt or sitemap.xml, but aimed at AI consumption rather than crawling permissions or URL discovery.
The goal is to reduce the work an AI system does to understand a site and to steer it toward the content the owner wants cited. A companion /llms-full.txt can inline expanded content for models that accept it. Adoption is voluntary and evolving, but for organizations that want to be accurately represented in AI answers, publishing one is a low-cost, high-signal move.
