Short answer
llms.txt is a proposed plain-text file at a site’s root that offers language models a curated map of its most useful pages, in Markdown, without navigation and boilerplate. It is a community convention rather than a standard backed by the major search engines, and adoption by large AI crawlers remains limited.
Also called: LLMs.txt
The file is cheap to publish and genuinely useful for agentic coding and research tools that read a site directly. Treated as a citation lever, however, it is oversold: Google has said it is unnecessary, and the major AI crawlers rarely request it. A site that publishes one and changes nothing else should not expect a visibility change.
Where it does earn its keep is as curation. A hand-written llms.txt states, in your own words, what the business does and which pages matter — useful whenever a tool reads it, and a low-cost forcing function for describing your own offering precisely. The failure mode is publishing an auto-generated default that selects pages arbitrarily and omits half the content worth reading.
Common questions
Does llms.txt actually do anything?
It is read by some agentic developer and research tools, and largely ignored by the big AI crawlers. Publish one because it is inexpensive and forces a clear description of your site, not because it is expected to produce citations on its own.
Related terms
Answer Engine Optimization (AEO)
Answer engine optimization is the practice of structuring content so that AI answer engines — ChatGPT, Perplexity, Claude, Copilot, Google AI Overviews — cite it inside their generated answers.
Generative Engine Optimization (GEO)
Generative engine optimization is the practice of increasing how often a brand or page is used as a source by generative AI systems.
robots.txt
robots.txt is a file at a site’s root that tells automated crawlers which paths they may request.
Reading definitions because you are scoping a project? Skip ahead and just ask.