llms.txt: what it is, and does it help AI recommend you?

The short answer

llms.txt is a Markdown file at yoursite.com/llms.txt that lists a site's most useful pages for AI tools. Google says it doesn't use it, no major AI assistant says it does, and studies of 300,000 sites found no link between having one and being cited by AI. It's harmless, and useful for software documentation, but a local business doesn't need one.

What llms.txt is

llms.txt is a plain text file in Markdown, placed at the root of a website (yoursite.com/llms.txt). It was proposed in September 2024 by Jeremy Howard as a way “to provide LLM-friendly content”: a short summary of the site and a list of its most useful pages, so an AI tool can find the right page without reading the whole site.

The format is simple. Only the first line is required:

A heading with the site's name (the only required part).
A one-paragraph summary, written as a quote.
Lists of links under headings, each with a short note. A section called “Optional” holds pages an AI tool can skip.

The proposal also suggests a clean Markdown copy of each useful page (for example /pricing.md). A related file you may see, llms-full.txt, puts the full text of a site's documentation in one file. That one is a common habit of documentation tools, not part of the proposal.

llms.txt vs robots.txt

The names are similar, but the jobs aren't:

robots.txt compared with llms.txt
	robots.txt	llms.txt
What it does	Tells crawlers what they may and may not fetch	Suggests which pages are worth reading
Who follows it	Google, Bing and the AI crawlers from OpenAI, Anthropic and Perplexity	No major AI search engine has said it uses it
If it's wrong	You can disappear from search and AI answers	Nothing happens
Do you need it?	Yes, and it should let AI search crawlers in	No

If you only check one file, check robots.txt. A wrong rule there can hide you from ChatGPT; a missing llms.txt can't. Our guide to AI crawlers explains which ones to allow.

Does ChatGPT or Google read llms.txt?

Google: no. Google's guide to its AI features says you don't need “AI text files” to appear in Google Search, that Search itself doesn't use them, and that an llms.txt file will neither harm nor help your visibility there. John Mueller of Google put it more bluntly in June 2026: a file a website writes about itself “can't” be trusted to tell good sites from bad ones. He saw one possible use: helping an automated tool that is already on your site find its way around.

OpenAI, Anthropic and Perplexity: their crawler documentation talks about robots.txt and nothing else. None of them says llms.txt affects whether a site is found or cited. All of them publish llms.txt files for their own developer documentation, which is what the format was designed for, and their crawlers do fetch the file now and then. But fetching a file isn't the same as using it to decide who gets recommended.

What the evidence says

Two large studies have looked at it, from different angles:

Studies of llms.txt and AI visibility
Study	What they looked at	What they found
SE Ranking, Nov 2025	About 300,000 domains: does having llms.txt go with being cited by AI?	10% of the domains had one. No link with how often AI cited them; the file made their prediction model slightly worse
Ahrefs, Jun 2026	137,210 sites: does anything read the file?	28% had one (a tech-savvy sample). 97% of those files got no requests at all in a month, from people or bots

Neither study proves llms.txt can never matter. They show that today it isn't read much, and that sites with one aren't cited more. If an AI company announces it uses the file, that changes, and we'll update this page.

Who should have one

Software and API documentation. This is where llms.txt is useful today: developers point AI coding tools at it so they read the right docs. In the Ahrefs data, AI coding tools were among the few that fetched it.
Large content sites that want to try it. It costs little, and it's harmless.
Local businesses: not needed. A restaurant, clinic or law firm gets nothing measurable from it. Your time is better spent on the steps below.

What gets a local business into AI answers instead

Let AI search crawlers in. robots.txt and your firewall or CDN must allow OAI-SearchBot, Claude-SearchBot, PerplexityBot, Googlebot and Bingbot.
Say what you do and where, as text. Services, city, neighbourhood, prices and opening hours on your own pages, not only in images or PDFs.
Add LocalBusiness structured data. Or a specific type like Restaurant or Dentist, with your name, address, phone and hours.
Keep your listings complete and the same everywhere. Google Business Profile, Bing Places, Apple Business Connect and the directories in your industry.
Be on the sites AI cites for your city. Guides, review sites and local press. AI often names businesses from these, not from their own websites.

The rest is in our plain-English guide to GEO.

If you want one anyway: a 10-minute llms.txt

It does no harm, so if you'd like one, keep it short and factual. Save it as llms.txt in the root folder of your site, next to robots.txt. A clinic's file could look like this:

llms.txt: example for a dental clinic

# Example Dental

> Family and cosmetic dental clinic in central Manchester. Same-day emergency
> appointments, Monday to Saturday. NHS and private patients.

## Services
- [Emergency dentist](https://www.example.com/emergency): same-day appointments, prices
- [Dental implants](https://www.example.com/implants): from £2,200, free consultation
- [Teeth whitening](https://www.example.com/whitening): in-clinic and take-home

## Visit us
- [Contact and opening hours](https://www.example.com/contact): 1 Example Street, Manchester

## Optional
- [About our dentists](https://www.example.com/team)

Write the same facts that are on your pages, and update the file when they change. A file that says something your website doesn't won't help, and an outdated one can only confuse. We publish one on stuckup.app too, mostly so people who look for it find it.

Questions

Does llms.txt help SEO?

No. Google says its Search doesn't use llms.txt, and that the file will neither harm nor help your visibility in Google Search, including AI Overviews and AI Mode.

Does ChatGPT read llms.txt?

OpenAI hasn't said it uses llms.txt to find or cite websites; its crawler documentation only covers robots.txt. Its crawlers fetch llms.txt files now and then, but studies found no link between having one and being cited.

Can llms.txt hurt my website?

No. Google says it neither harms nor helps. The only risk is an outdated file that says something different from your website, so keep it short and update it with your pages.

What is llms-full.txt?

A single file with the full text of a site's documentation, made by many documentation tools. It isn't part of the llms.txt proposal, and it's only useful for documentation sites.

Is llms.txt the same as robots.txt?

No. robots.txt tells crawlers what they may fetch, and search engines and AI crawlers follow it. llms.txt only suggests pages to read, and no major AI search engine has said it uses it.

Sources

Narek Nikoghosyan · Founder and CEO
Narek Nikoghosyan is the founder and CEO of stuckUP, the AI visibility checker for local businesses.