llms.txt
A proposed root file curating a site's key content for AI systems. Adoption is limited; it does not replace being indexed.
llms.txt is a proposed convention: a Markdown file at the root of a site (/llms.txt) that offers large language models a curated, human-readable map of the site's most important content — what the site is, and links to the pages worth reading — so an AI can find the signal without wading through navigation and boilerplate.
It's worth understanding what it is and isn't. It's a community proposal, not an established standard, and as of now major AI providers have not committed to reading it, so its practical impact is limited and uncertain. It is not the same as robots.txt (which governs crawling) and it does not grant or block anything.
The honest framing: publishing a clean llms.txt is low-cost and can't hurt, but it is not a shortcut to AI visibility. Being crawlable, indexed and genuinely authoritative on Google and Bing remains the thing that actually determines whether AI systems can retrieve and cite you. Treat llms.txt as a nice-to-have curation layer on top of fundamentals, not a substitute for them.
Related terms
- AEO (Answer Engine Optimization) — Structuring content so answer engines can lift a direct answer from it — for featured snippets, voice results and AI answers.
- GEO (Generative Engine Optimization) — Improving how often generative AI answers (ChatGPT, Perplexity, Gemini) cite your site. Visibility is measured, not bought.
- Indexing — Storing a crawled page in a search engine's index so it can appear in results. Crawled is not the same as indexed.
- robots.txt — A file at the site root that tells crawlers which paths they may or may not fetch. It controls crawling, not indexing.