No. An llms.txt file is harmless and cheap, but there is no evidence it is read. Ahrefs analysed all 137,210 domains in its web analytics that received traffic in May 2026: about 28% had published an llms.txt file, and 97% of those files received no requests at all during the month. The few that were fetched were requested by SEO audit tools rather than by AI systems (Ahrefs, June 2026). What AI agents actually read is your ordinary HTML.
Why the Advice Persists
llms.txt was proposed as a way to give language models a curated summary of a site, and it spread quickly because it is easy to publish and easy to recommend. Ahrefs notes that its 28% adoption figure is an upper bound, since its customers skew technical and SEO-aware. Adoption has run far ahead of usage: the file exists on a large share of sites and is consulted on almost none.
What Agents Read Instead
A buyer's procurement agent searches the web for companies matching a brief, reads the pages it finds, and looks for an address it can send a request to. The things that decide whether it can act are:
- A plain-text email address on a page that loads without JavaScript. An agent that cannot reach you drops you from the shortlist.
- Capabilities, materials, regions and certifications as HTML text, in the vocabulary buyers use. PDFs and images are not read.
- Pages a crawler can fetch: server-rendered or prerendered content, no login wall, and a robots.txt that does not block AI user agents.
- Organization schema with name, address, contact point and `sameAs` links, which clarifies which entity you are.
None of these depend on llms.txt, and all of them outrank it. The full checklist and the reasoning are in the guide to how to be found by AI buying agents.
Should You Publish One Anyway?
If it takes ten minutes, yes, as hygiene. Buyer24 publishes one. Do not let it displace the contact page or the capabilities text on your to-do list, and do not expect it to change what agents find. If it is the only thing you do for AI discoverability, you have done approximately nothing.
FAQ
What is llms.txt?
A plain-text file at the root of a website that summarizes the site and lists key pages for language models, modelled loosely on robots.txt. It is a proposed convention, not a standard adopted by the major AI systems, and current measurement shows it is rarely requested.
Does llms.txt hurt anything?
No. It is a static file that costs nothing to serve. The only cost is opportunity: time spent writing it is time not spent putting a plain-text email address and readable capabilities on the pages that agents actually fetch.
What about schema markup instead?
Schema.org markup is more useful than llms.txt, for a narrower reason. Organization schema clarifies your identity and contact point; Product or Service schema helps if you have stable SKUs. Ignore claims that markup makes you several times more likely to be selected by agents; those figures circulate without a traceable source.
People also search for: