llms.txt — A proposed standard: a markdown file at the root of a domain listing the pages that matter, intended to help AI systems find the useful content.
What it is
A plain markdown file served at /llms.txt, listing a site’s most important URLs with a one-line description of each. It was proposed by Jeremy Howard of Answer.AI in September 2024, by analogy with robots.txt — but where robots.txt says what not to crawl, llms.txt says what is worth reading.
The honest status
Adoption is uneven and largely unverified. Some AI vendors are reported to read it; others have said nothing. Google has not endorsed it, and has signalled that its preferred signals remain robots.txt, sitemaps and structured data.
Treat anyone selling llms.txt as a ranking factor with suspicion. There is no public evidence that it moves visibility on any major surface, and claiming otherwise is the kind of thing that will age badly.
Why publish one anyway
It costs nothing, the downside is zero, and the exercise itself is useful — writing it forces you to decide which of your pages actually matter and describe each in one sentence. Most sites cannot do that without arguing about it first, which tells you something.
Ours is at foundfirstsearch.com/llms.txt, generated automatically from the live pages so it cannot go stale.
Related: GEO, entity resolution.