What Is llms.txt and Should Your Store Have One?

llms.txt is a proposed plain text file, published at yourstore.com/llms.txt, that hands an AI model a short curated map of your site: your most useful pages, in reading order, with a line explaining each one. It is a suggestion, not a standard. No major AI provider has confirmed that it changes anything, and Google has said publicly that it does not use it.

The verdict up front: publishing one costs an hour and cannot hurt you, and it is not a visibility strategy. If you are choosing between writing an llms.txt file and fixing the content on the pages you already have, fix the content.

What the file actually is

The idea was proposed in late 2024 by Jeremy Howard of Answer.AI. Models work from a limited amount of text at a time, and a modern store page is mostly navigation, scripts, and markup wrapped around a few hundred words that matter. llms.txt is a markdown file that skips all of that: a title, a sentence about the business, and then linked lists of the pages worth reading, each with a short description.

Some sites also publish llms-full.txt, which contains the actual text of those pages rather than links to them. Both files live at the root of the domain, like robots.txt, and both are plain text you can open in a browser.

What it is not

This is where most explanations go wrong, so it is worth being blunt. llms.txt is not an access control file. It does not block AI systems, it does not grant them permission, and it does not protect anything. Publishing one cannot keep a model away from your customer data, because it was never a gate.

The file that governs crawling is robots.txt, which is honored by the crawlers that choose to honor it, including the ones operated by the major AI companies. If your goal is to keep a crawler out, that is the file to edit, and the relevant question is which user agents you name. llms.txt answers the opposite question: for the systems you do want reading your site, here is the good stuff, described clearly.

Worth saying plainly: anything on a public product page is public. Nothing in either file makes a page private.

Who reads it today

Nobody has confirmed that they do, at least not in a way that affects what an assistant tells a shopper. Google has stated that it does not use llms.txt, with one of its search advocates comparing the idea to the old keywords meta tag, a file where a site owner describes itself and hopes the description is believed. OpenAI's crawler documentation describes robots.txt behavior and does not mention llms.txt at all.

The sites that do publish llms.txt are overwhelmingly developer documentation sites, including the AI companies themselves and infrastructure vendors. That tells you what the file is good at right now: helping a coding assistant read technical docs on request. It is a minority practice on the open web, and it is not concentrated in ecommerce.

So should your store publish one?

Publish one if all three of these are true. Your site has a handful of genuinely useful pages that are hard to find from the homepage, like buying guides and sizing information. You can write the file in an hour. And you can accept that the return might be exactly zero.

Skip it if writing it means not doing something else this week. A store that gets recommended by an assistant almost always got there by having crawlable pages that answer a buying question specifically, with real products in them. No index file substitutes for that, and no index file compensates for its absence.

The honest framing is that llms.txt is a cheap bet on a convention that might get adopted. Bets like that are fine as long as you price them correctly.

What a store's llms.txt looks like

The format is markdown, and the whole file should fit on one screen. Lead with what you sell, then link the pages that answer questions, each with a description that says what the reader will get.

# Northwind Coffee

      > Small-batch coffee roaster in Portland, Oregon. Single origin beans,
      > subscriptions, and brewing gear, shipped within one business day.

      ## Buying guides
      - [Choosing a grind size](https://example.com/blogs/news/grind-size):
      Which grind suits which brewer, with a chart and our recommendations.
      - [Light roast vs dark roast](https://example.com/blogs/news/roast-levels):
      What changes in flavor, caffeine, and brew time.

      ## Shop
      - [Single origin beans](https://example.com/collections/single-origin):
      Current lots with tasting notes, origin, and roast dates.
      - [Brewing gear](https://example.com/collections/gear): Grinders, kettles,
      and filters we actually use.

      ## Policies
      - [Shipping and returns](https://example.com/policies/shipping): Ships in
      one business day, 30 day returns, free over $45.

Upload it in Shopify under Content, then Files, and make sure it resolves at your domain followed by /llms.txt. Then leave it alone until your catalog or your guides change. A stale file that describes last season is worse than no file, because the one thing you can count on is that anything reading it will believe it.

What to do instead, or first

The work that moves the needle is unglamorous. Make sure your key pages render their content without JavaScript, because a crawler that sees an empty shell learns nothing. Answer one buying question per page, in the first paragraph, in plain language. Put your actual products in those answers, with the specifics a shopper would compare. Keep your product data accurate, since an assistant reading a stale price will quote a stale price.

We publish an llms.txt file for riten.ai, generated automatically when the site builds, and we would still tell you the same thing: it is a courtesy to whatever reads it, not the reason anyone gets recommended.

FAQs

Is llms.txt the same as robots.txt?

No. robots.txt tells crawlers which paths they may fetch, and the major AI crawlers honor it. llms.txt makes no rules at all. It is a curated index for systems that already have access, so it cannot block a crawler or protect a page.

Does Google use llms.txt?

No. Google has said publicly that it does not support the file and is not planning to, comparing the idea to the old keywords meta tag. Appearing in Google's AI features depends on being indexed and eligible for a normal search snippet, not on any additional file.

Can publishing llms.txt hurt my store?

Only through inaccuracy. The file is inert, so the realistic risk is describing products, prices, or policies you no longer offer and having a model repeat it. Keep it short, keep it current, or do not publish it.