Straight answers
llms.txt: what it does, and what it doesn't.
You may have been told that adding an llms.txt file will get your store into AI answers. Here is what the evidence says, checked on 6 October 2026. We ship one ourselves, so we have no reason to talk it down, and no reason to oversell it.
What is llms.txt?
A Markdown file at the root of a site, for example yourstore.com/llms.txt, that lists the pages a language model should read first. Jeremy Howard of Answer.AI proposed it on 3 September 2024, with the format at llmstxt.org.
It is a proposal, not a standard, and it is not an access rule. Whether a crawler may read your store is decided by robots.txt and by your server or CDN, not by this file.
Does Google use it?
No. Google's own guidance on generative AI in Search says so directly:
"You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search (including its generative AI capabilities), as Google Search itself doesn't use them."Google Search Central, updated 10 July 2026
Do ChatGPT, Claude and Perplexity read it?
Rarely, by the best data available. Ahrefs looked at 137,210 sites in May 2026. 28% published an llms.txt file, and 97% of those files received no requests at all that month. Where a site had no file, no AI bot came looking for one. (Ahrefs, June 2026.)
Anthropic and OpenAI both publish an llms.txt for their own developer documentation. We found no public statement from OpenAI, Anthropic or Perplexity that their assistants use other sites' files to decide what to recommend.
So who is it for?
Mostly for tools that read documentation on a developer's behalf, such as coding assistants. That is a real use. It is just not a shopper asking an assistant where to buy something.
Should your store have one?
If someone offers it for an hour of work, fine. It does no harm, and we ship one ourselves (here it is). What we will not do is sell it as the reason you will start being recommended.
What decides that is more ordinary and easier to check:
- Whether AI crawlers can reach you at all. In our measurement of 810 US stores, 61% refused a request identifying itself as GPTBot. Test yours in ten seconds.
- Whether your product facts are in the raw HTML. Price and availability that appear only after JavaScript runs are invisible to most crawlers.
- Whether those facts are machine-readable.
Productstructured data states them without guesswork.
The two-minute self-check covers all of these.
Short answers
Will an llms.txt file get my store recommended by ChatGPT?
There is no evidence that it will. Google says its search ignores the file, and in a study of 137,210 sites 97% of llms.txt files were never requested in a month.
Can llms.txt block AI crawlers?
No. It is a reading list, not an access rule. Blocking or allowing crawlers is done in robots.txt and in your server or CDN settings.
Should I remove the llms.txt I already have?
No need. It costs nothing to keep and does no harm. Just do not count it as the work.