What llms.txt actually is
A short, practical guide — not a claim that adding this file will get you cited by an AI system.
What it is
llms.txt is a proposed convention — placed at the root of a domain, like /llms.txt — for giving AI systems a concise, machine-readable summary of a site: what it is, its key pages, and how the owner would like it referenced. It's written in plain Markdown, similar in spirit to robots.txt or sitemap.xml, but aimed at language models reading for context rather than crawlers indexing pages.
What it isn't
It isn't a standard any AI provider has committed to reading or honoring, and it isn't a ranking or citation signal in the way a sitemap is a crawling signal. Treat it as one readiness signal among several — alongside AI-bot access in robots.txt, structured data, and whether your content is actually extractable from server-rendered HTML — not a guarantee of anything.
What to put in one
- A one-paragraph description of what the site/company is and does.
- Links to the pages that best represent the site — docs, pricing, key product pages.
- A short note on how you'd like the site attributed or described, if that matters to you.
- Keep it short. It's a summary, not a full site export.
A minimal example
# Acme Widgets
Acme makes widgets for small manufacturers.
## Key pages
- Product: https://acme.example/product
- Docs: https://acme.example/docs
Where this fits in Seov's GEO score
Seov checks whether /llms.txt exists and is non-empty, as one signal in the GEO lane's extractability category — worth doing, low effort, but never weighted as if it were the deciding factor.