llms.txt is a plain-text file at the root of a website that lists the site’s most important pages for AI tools to read. It is a proposed convention, not a requirement, and it does not make an answer engine cite you. Treat it as housekeeping that costs little, and measure results with tracked questions instead.
The file was proposed in 2024 as a clean, short index of a site’s canonical content, written in Markdown. The sections below explain what it contains, what the evidence says about its use, and how a small business can check its own file.
Why was llms.txt proposed?
llms.txt was proposed because web pages are written for browsers, with menus, scripts and layout code around the text, while an AI tool needs the useful content with little else. The proposal suggests a short Markdown file that names the pages that matter and describes each in a sentence, so a tool can find them without reading the whole site.
The idea is similar to a sitemap, with two differences. A sitemap lists every indexable URL for search crawlers, and llms.txt lists a short, curated set of pages with descriptions written for language models. Neither file decides how an engine ranks or cites a page.
What does an llms.txt file contain?
An llms.txt file contains a site name, a one-line summary and a list of links to key pages, each with a short description. It sits at yourdomain.com/llms.txt and is written in Markdown so that both people and software can read it. A typical small business file links to the About, Services, Pricing, Contact and policy pages.
The file does not hold new content. It points to pages that already exist, which is why the pages it lists need to be accurate and current.
Do AI engines read llms.txt?
Engine support is unconfirmed. In June 2025, Google’s John Mueller wrote that no AI system currently uses llms.txt, and that chatbots fetch pages but none fetch the llms.txt file. Some vendors and plugin makers say otherwise, and engines do not publish full details of how they select sources, so treat this as a statement about that date, not a permanent fact.
This is the reason to avoid treating the file as a visibility tactic. If an engine does not read it, the file changes nothing about its answers.
If your goal is to be named in an answer, the practical route is the same as before: pages that state facts clearly, listings that agree with them, and a record of what the engines return.
What does llms.txt not do?
llms.txt does not control, block or guarantee anything. Four limits are worth knowing:
- It is not robots.txt. It cannot stop a crawler from reading a page and it does not set crawl rules.
- It does not guarantee citations. No file on your site decides what ChatGPT, Gemini or Google AI Overviews say.
- It does not replace schema markup. JSON-LD describes your organization and pages to search engines in a form they already support.
- It does not fix wrong facts. If a directory lists an old address, the file will not correct it.
Accurate pages, consistent business details and corrections made at the source do the real work.
Free audit
See where your business shows up in AI answers
We run ten of your buying questions through ChatGPT, Gemini and Google AI Overviews and show you where you stand. A person replies within 24 hours, Monday to Friday. You keep the audit either way.
How can a small business check its llms.txt?
Open yourdomain.com/llms.txt in a browser and check four things: the file loads as plain text, the links are live, the pages listed are the ones you want described, and no private or outdated page appears. Many WordPress SEO plugins can generate the file. AvantScale’s own site generates one through Yoast SEO, with the page list set by hand so it includes the Services, Pricing, About, Contact and Locations pages.
Review the file whenever you change your main pages. A file that lists a page you have since removed points readers to an error.
What should you measure instead?
Measure what engines say, not whether a file exists. Pick the questions your customers ask, run them on ChatGPT, Gemini and Google AI Overviews on the same day, and log whether your business is named, who is named instead and whether any fact is wrong. Repeat the test monthly with the same wording.
Keep the test set small and stable. Ten well-chosen buying questions, run the same way each month, give a cleaner comparison than a long list that changes every time.
That log shows whether a change, including adding llms.txt, coincides with a difference in answers. It shows correlation on one set of questions, not proof that the change caused it. The consulting and audits page explains how a baseline test is run, and current prices are on the Pricing page.
llms.txt: FAQ
Do I need an llms.txt file?
No engine requires one. If your SEO plugin generates it at no cost, publishing a correct file is reasonable housekeeping. Do not spend significant time or budget on it, and do not expect it to change what any engine says.
Is llms.txt the same as robots.txt?
No. robots.txt gives crawlers rules about which parts of a site they may access. llms.txt is a reading list of key pages, and it cannot allow or block anything.
How do I know whether llms.txt helped?
Test the same set of buying questions before and after, on the same engines, with the same location setting, and compare the logged answers. Answers vary between sessions, so a single change in one answer is not evidence.
Measure what engines actually say about you
The free audit tests 10 of your buying questions on ChatGPT, Gemini and Google AI Overviews and records what each one says, with a screenshot of every answer. Get your free audit.
