Glossary

llms.txt is a courtesy note for AI crawlers, not an AI Overviews ticket

llms.txt is a plaintext file placed at a website's root, listing its key pages in a simple Markdown format for AI systems that choose to read it. Google's own guidance says it plays no role in AI Overviews or AI Mode. A handful of other systems, including Claude, Perplexity, and various coding agents, do document support for it, so the file is real but far narrower than its reputation.

Definition

llms.txt is a proposed convention, not a web standard, in which a site publishes a short Markdown file at its root listing its most important pages and a brief description of each, intended to give AI systems a quick index of the site if they choose to fetch it. Adoption among the systems that actually honour it remains limited and system-specific.

What the file actually contains

A typical llms.txt is short: a one-line description of the site, then a Markdown list of links to its most important pages, each with a short note on what that page covers. It sits at the domain root, the same location as robots.txt, and it is meant to be read by a system rather than a person.

It is not schema markup, not structured data in any format search engines already parse, and not a replacement for a sitemap. It is a separate, voluntary file that only matters to the specific systems that have chosen to look for it.

Which systems read it, and which explicitly do not

Google's Search Central AI Optimization Guide, published 15 May 2026, lists llms.txt alongside Markdown alternates and artificial content chunking as things that are not ranking or citation factors for Search, AI Overviews, or AI Mode. Googlebot gives the file no special processing, and Google engineers have said directly that Google does not support it and has no plans to.

That is guidance for Google's own products specifically, not a statement that no system uses the file. Claude, Perplexity, and a number of AI coding agents do document support for reading llms.txt when it is present. The honest position is narrower than either extreme: some systems read it, the largest one by search volume has said plainly that it does not.

What actual adoption looks like

Ahrefs analysed server logs across 137,000 domains in May 2026 and found that 97% of published llms.txt files received zero requests in that period. Of the small share that did get traffic, most of it came from bots unrelated to live AI retrieval, SEO audit tools, general web crawlers, and tech-profiling services, while AI retrieval bots tied to ChatGPT and Perplexity accounted for only about 1% of the requests logged.

97%

of published llms.txt files received zero server requests, measured across 137,000 domains in May 2026.

Ahrefs, llms.txt study

What earns AI citations instead

The systems that reliably read and cite a page are the ones parsing its actual HTML: clear headings, semantic structure, and schema that names the business, its category, and its facts. That is the layer AI-readable design actually operates on, and it works regardless of whether a given crawler ever checks for an llms.txt file.

Publishing an llms.txt file is a low-cost addition for the systems that do read it. It is not a substitute for that structural work, and treating it as an AI-visibility strategy on its own overstates what the file does.

A courtesy, not a lever

The file is cheap to publish and easy to overrate. What earns a citation is the structure inside the page.

Common questions

Does llms.txt help a site rank in Google AI Overviews?

No. Google's own AI Optimization Guide states plainly that llms.txt is not a ranking or citation factor for AI Overviews or AI Mode, and that Googlebot gives the file no special processing. Normal SEO fundamentals, crawlable pages and clear structure, are what feed those features.

So is llms.txt useless?

Not entirely, just narrower than the pitch. Some systems that explicitly document support for it, including Claude, Perplexity, and various coding agents, do read it. It is a cheap courtesy to those systems, not a general AI-visibility win.

Which businesses actually benefit from adding one?

Mostly sites with AI coding agents or AI-native tools as part of their audience, where the systems reading the file are the ones known to honour it. For a typical business site chasing citations in ChatGPT, Perplexity, or Google AI Overviews, the file is not where the effort belongs.

What should a business do instead of relying on llms.txt?

Put the structure inside the HTML itself: clear headings, semantic markup, and schema that names the business, its category, and its facts. That is what crawlers actually parse, and it works whether or not a given system reads llms.txt at all.

See what AI systems can actually read on your site before adding another file to it.

Get a visibility audit or talk to the studio

llms.txt is a courtesy note for AI crawlers, not an AI Overviews ticket