Does a Cafe Website Need an LLMs Text File?

Thomas Garner

Oct 7

10 min read

Someone forwarded you a screenshot, an SEO newsletter, or a setting inside your website builder that mentions llms.txt, and now you are wondering if you are missing a step that is quietly costing you customers. I run the company side of Hillcane, the part that spends its time reading documentation so owners do not have to, and I want to give you a straight answer here rather than a vague one.

Short answer: llms.txt is a proposal, written up by developer Jeremy Howard on the site llmstxt.org, meant to help an AI assistant find and use a website's content more efficiently. It is not a Google ranking signal, it is not a requirement from any major search or AI vendor, and as of the documentation I reviewed on September 21, 2026, none of the major assistant vendors' own crawler or search documentation confirms that their systems read or require the file. For a cafe, that makes it a low effort extra worth adding once your real pages exist, never a shortcut that replaces them.

This post walks through what the file actually is, what a builder like Wix does and does not provide around it, what a sensible version for a cafe should contain, and where the real limits sit so you spend the right amount of time on this, which is not very much.

Definition

llms.txt is a plain text file, placed at the root of a website (yourdomain.com/llms.txt), written in a simple Markdown format meant to summarize a site for an AI system in a compact way. The proposal was published by Jeremy Howard on the site llmstxt.org, which is still the place to read its current wording.

The idea behind it is straightforward: a normal webpage is built for a human to read, with navigation, images and styling, and that structure can be inefficient for an AI system trying to extract the useful content quickly. An llms.txt file is meant to offer that same content in a stripped down, easy to parse format, similar in spirit to how a sitemap.xml file helps a search engine find your pages, or how robots.txt tells a crawler what it can and cannot access.

The important distinction is that llms.txt is a proposal, not a standard that any major platform has agreed to support. It lives in the same category as an idea a developer put forward that some tools have experimented with, not in the category of something Google Search Central, OpenAI, Anthropic or Perplexity has published documentation requiring or confirming their systems read. That difference matters a great deal for how much effort you should spend on it.

Compare it to robots.txt, which has been a real, decades old, universally respected standard that essentially every crawler on the internet honors, or to a sitemap, which Google Search Central documentation explicitly describes as a way to help Google find your pages. llms.txt has neither that history nor that documented support yet. It might grow into wider use. It has not, based on the documentation available as of this writing, become that yet.

It helps to be precise about what kind of thing a proposal is on the internet. Plenty of good ideas have been floated as a simple text file convention over the years, and a small number of them, robots.txt among them, eventually became something every major crawler agreed to respect. Most proposed conventions never reach that point, and they simply become a niche practice adopted by a handful of enthusiastic sites without ever becoming something a search or AI vendor documents as required or even read. llms.txt is currently in that earlier, unproven stage, and a cafe owner deciding how much time to spend on it should weigh that status honestly rather than treating early enthusiasm online as confirmation.

Wix Support

If your cafe's site runs on Wix, the plain answer is that Wix's Help Center documents its sitemap and its robots.txt handling, and the help pages reviewed for this post do not describe a built in llms.txt setting in the same way. So this post describes only what those pages do cover, rather than a feature nobody has documented.

What I can say with confidence, because I checked it directly on September 21, 2026, is that hillcane.co, Hillcane's own site, has a live llms.txt file at the root of the domain. It opens with a plain description of the business, what a marketing studio for independent coffee shops does, and it functions as a small, readable summary of the site, distinct from a normal blog post page.

If your platform does not offer a dedicated setting for this file, the practical route is a manually created text file added at your site's root, following the same simple Markdown format the proposal describes: a short title, a one line summary, and then a handful of links to the pages that matter most. This is not a complicated build. It is closer in effort to writing a short bio than to building a page.

Whatever your platform, treat any claim that it automatically handles llms.txt for you with the same skepticism you would apply to any other newly popular setting. Confirm it directly on your own live site before assuming it is there, the same way you should periodically check that your sitemap and robots.txt are behaving the way you expect.

If you use Squarespace or Shopify instead of Wix, the same principle applies: check the actual root of your domain rather than trusting a plugin description or a forum post claiming automatic support. Platform features around a newly popular file convention tend to arrive unevenly, one platform first, others later, and the only reliable way to know what your own site currently does is to look at it directly.

What to Include

A sensible cafe version of this file is short, under twenty lines, and mirrors the structure the proposal describes: a title, a one paragraph description, and a short list of links to the pages an assistant would most want to point to. Keep it plain text, no styling, no marketing flourish.

Start with your cafe's name as a heading, followed by one paragraph describing what the shop is, where it is, and what makes it worth mentioning, written the way you would describe the place to someone who has never heard of it. Then list direct links to your hours and location page, your menu, your story or about page, and your contact page. That is the whole file.

Before you publish it, click every link in the file and confirm each one loads correctly and returns a normal page, not a broken link or a redirect chain. A file meant to help an assistant find your real content is worthless if the links inside it are wrong, and nobody is likely to notice a broken link buried in a text file the way they would notice one on your homepage.

Do not use this file as a place to cram keywords, a list of every menu item, or promotional language. The whole value of the file, to the extent it has value at all, is as a plain, accurate summary. Treat it the way you would treat a short bio you hand someone before an interview: accurate, brief, and free of sales copy.

Once the file is live, revisit it whenever your core pages change meaningfully, such as a new menu structure, a new hours page address, or a change of location. A stale llms.txt pointing to old links or describing a version of the shop that no longer exists is worse than no file at all, since it hands out inaccurate information dressed up as a summary. Treat it as a small piece of your site to check during any larger update, not a set it and forget it task.

Limits

Here is the honest limit: as of the vendor documentation search conducted September 21, 2026, none of the major AI vendors' publicly available crawler or search documentation confirms that their systems read or require an llms.txt file. That does not mean no assistant ever looks at one. It means there is no documented, confirmed statement from a major vendor saying so, and a post like this should not claim more certainty than the documentation supports.

Hillcane's own site is a useful, honest example of why this file cannot be evaluated as a magic switch. The site launched September 9, 2026, and in its first twelve days saw roughly 2,000 AI bot hits out of about 7,738 total requests, alongside the first Google impressions on day one to two, about 124 impressions per day by day eight, first organic clicks on day seven, and the first ChatGPT tagged visits around days eight to nine. That is a real, small, dated case, not a study, and no export on file isolates whether any of that traffic had anything to do with the llms.txt file specifically versus the site's actual pages, its schema, or its sitemap. Do not read more into that number than it supports.

The clearest way to think about this file is as a nice to have addition to a foundation that already exists, never a substitute for it. Real pages with real hours, a real menu written in actual text rather than only a photo or PDF, a working sitemap, and clean crawling access matter far more, and each of those has its own dedicated coverage elsewhere on this site rather than being repeated here. If you have not read through what a crawler does on a brand new site, or how to decide which crawlers to allow, those pieces cover ground this post intentionally leaves alone.

If you have ten minutes and your real pages are already in good shape, adding a short llms.txt file costs you almost nothing and cannot hurt. If your hours page is out of date, your menu only exists as an image, or your sitemap has never been checked, spend those ten minutes there first. That is where the actual gain sits, and no proposal file changes that order of priorities.

It is worth saying why this order matters so much for a small business specifically. A larger company might have someone whose entire job is testing new conventions like this one, with time to spare either way. A cafe owner does not have that spare time, so every minute spent on an unproven convention is a minute not spent on hours, menu text, schema markup, or a sitemap, each of which has documented, confirmed support from the tools that send customers your way. Order of operations is the whole argument here, not a blanket dismissal of the file itself, and it is an argument worth repeating the next time a newsletter tells you a small text file is the one thing standing between your cafe and being cited by an assistant.

Frequently Asked Questions

Should I add an llms.txt file to my cafe website?

It is a reasonable, low effort addition once your real pages, hours, menu and contact information are solid, but it should never come before them. As of September 21, 2026, no major AI vendor's documentation confirms its systems require or read this file, so treat it as a small extra, not a fix for a thin site.

Does Wix automatically create an llms.txt file for my site?

Wix's Help Center documents how Wix handles sitemaps and robots.txt, but the help pages reviewed for this post do not describe an automatic llms.txt setting in the same way. Check your own live site directly at yourdomain.com/llms.txt to see what, if anything, exists there. If nothing does, a short manually created text file at the root, following the proposal's simple Markdown format, does the job.

Do ChatGPT, Claude or Perplexity read llms.txt files?

As of the vendor documentation search on September 21, 2026, none of the publicly available crawler or search documentation from OpenAI, Anthropic, Google or Perplexity says their systems read or require this file. It is a proposed convention published on llmstxt.org, not a documented standard those systems have agreed to support, so treat any claim that an assistant depends on it with care.

What should a cafe's llms.txt file contain?

Keep it short, under twenty lines of plain Markdown: a heading with your cafe name, one plain paragraph describing what the shop is and where it is, and a handful of working links to your hours and location page, your menu, your story page and your contact page. No keyword stuffing and no menu dump; click every link before publishing to confirm it loads.

Will adding llms.txt help my cafe rank higher in Google?

No documentation reviewed states that this file affects Google ranking. It is unrelated to the Rich Results Test, Search Console, or Google's organic ranking systems, and Google Search Central does not list it alongside sitemaps and robots.txt. Treat it purely as a possible AI assistant convenience file, and spend ranking effort on hours, text menus, schema and a working sitemap instead.

Is llms.txt the same as robots.txt or a sitemap?

No. Robots.txt is a decades old standard that essentially every crawler honors, and Google Search Central documentation explicitly describes sitemaps as a way to help Google find your pages. llms.txt is a newer proposal from developer Jeremy Howard, published on llmstxt.org, without that same history or confirmed support from major crawlers. It might grow into wider use; it has not yet.

How long does it take to add an llms.txt file?

Well under an hour for most cafes. Writing the one paragraph summary and gathering four or five links, hours and location, menu, story and contact, takes a few minutes, and testing that each link loads correctly takes only a few more. It is a small task for a slow afternoon, not a project, and it is worth revisiting whenever those core pages change.

Does hillcane.co have an llms.txt file?

Yes. Checked directly on September 21, 2026, hillcane.co has a live llms.txt file at its root that opens with a plain description of the business. The site launched September 9, 2026 and saw roughly 2,000 AI bot hits out of about 7,738 requests in its first twelve days, but no export isolates the file's effect from the site's pages, schema or sitemap.

Should I skip real website pages to focus on llms.txt instead?

No. Real, complete pages matter far more than this file: current hours, an actual text menu rather than only a photo or PDF, a working sitemap and clean crawling access. If those foundations are not solid yet, spend your ten minutes there first. The llms.txt file is only worth adding once the real pages already exist, since its links have to point somewhere useful.

How can Hillcane help with AI search basics like this?

Hillcane helps cafe owners sort documented AI search practices from newsletter hype, and builds the foundational pages that matter first: hours, a text menu, schema and a working sitemap, with a short llms.txt file added afterward if you want one. Reach out at hillcane.co/contact or call (256) 384-2449 to talk through what your site needs before anything else.

Work with Hillcane

Someone mentioned an llms.txt file and now you are wondering if skipping it is costing you customers. It probably is not, and here is what the documentation actually says.

If you want a plain read on what your cafe website actually needs before chasing every new file or setting, reach Hillcane at hillcane.co/contact or call (256) 384-2449.