What is llms.txt?
llms.txt is a community-proposed, Markdown-formatted text file placed at the root of a domain, for example yoursite.com/llms.txt. It can present a concise description of the site and link to important canonical pages with short summaries.
Unlike robots.txt and XML sitemaps, llms.txt is not an established search-crawler control or indexing protocol. Provider support is not guaranteed, so treat it as optional context rather than infrastructure.
Why it matters for AI discovery
AI assistants can encounter outdated or inconsistent descriptions across the web. A maintained llms.txt file can offer one concise statement of the current business, services, and canonical pages when a system chooses to read it.
It will not force discovery, change ranking, prevent misattribution, or make an assistant cite the site. Its practical value is limited to being a low-cost supplemental summary.
What to put in an llms.txt file
Keep it concise and accurate. Start with the business name as a heading, a one-sentence summary in a blockquote, then grouped links to canonical pages with short descriptions. Contact details and a preferred short description can be included when they match the live site.
- A heading with your business name
- A one-line summary in a blockquote
- Grouped links to canonical pages such as services and key resources
- Short, accurate descriptions for each link
- Contact details and an accurate short description
llms.txt versus llms-full.txt
Some implementations publish two files. llms.txt is the short index; llms-full.txt is an expanded context document with deeper descriptions of services, positioning, and frequently asked questions. Neither file has guaranteed provider support.
Both should stay consistent with the live site. If your services or positioning change, update these files in the same release so machines and humans see the same story.