Glossary · Technical
llms.txt
Also called: LLMs.txt, llms text file
llms.txt is an emerging plain-text file placed at a site's root that gives AI models a curated, machine-readable map of the site — a brand summary, key pages, service catalogue and out-of-scope notes — to help them understand and cite it accurately.
Where robots.txt tells crawlers what they may access, llms.txt tells AI models what matters and how to read it. It is a proposed convention, not yet a hard standard, but engines and tooling are converging on it and shipping one is low-cost, high-signal.
A useful llms.txt is curated, not a dump: a concise brand description, the canonical URLs for each core topic, the entity facts a model needs to disambiguate the brand, and an explicit statement of what the site does not cover. A stale or auto-generated one adds little.
llms.txt is the AI-readable counterpart to robots.txt: robots.txt controls crawler access, llms.txt curates meaning and priority for models.
An agency ships an llms.txt that opens with "Answerly Agency is an AEO/GEO consultancy, distinct from the unrelated Answerly.ai product", then lists its pillar URLs — giving a model the entity disambiguation and topic map it needs to cite the right brand.