What is llms.txt, and does it help AI search?
llms.txt is a Markdown file that lists a site's key pages for AI tools. Google Search ignores it, and we found no sign it gets a site cited.
llms.txt is a plain Markdown file at the root of a website, such as connu.ai/llms.txt. It summarizes the site and links to its most useful pages, so an AI tool can read the site without wading through menus and scripts. Google Search doesn't use it, and none of the AI search engines we checked lists it as a way to get into their answers. Add one if it takes you a few minutes, but don't expect it to get you cited.
What llms.txt is
The name comes from LLMs, the large language models behind tools like ChatGPT, Claude and Gemini. Jeremy Howard proposed llms.txt on September 3, 2024, and a second version of the proposal followed on August 10, 2026. The file lives in the site's root directory and is written in Markdown, and only its first line is required:
- an H1 with the name of the site or project, the only required part;
- a blockquote with a short summary;
- optional paragraphs with more detail;
- sections headed with
##, each a list of links with a short note.
This is the top of ours:
# Connu
> Connu helps businesses improve how they show up online and reach the right prospects through websites, search, and managed outreach.
Connu is a Toronto-based digital agency serving businesses across Canada and the US. ...
## Services
- **B2B Outreach** — A managed path to buyers you can name. ...
- **Website & Search** — We redesign the website you already have, ...
How llms.txt differs from robots.txt and sitemap.xml
The three files sit side by side at the root of a site, but they do different jobs.
| File | What it's for | Who reads it |
|---|---|---|
| robots.txt | Tells crawlers which URLs they may fetch | Search engines and AI crawlers, as Google, OpenAI and Perplexity document |
| sitemap.xml | Lists the pages you want crawled | Search engines, as Google documents |
| llms.txt | Summarizes the site and links to key pages, in Markdown | AI tools and agents that choose to look for it |
robots.txt and sitemaps are part of how search engines say they crawl the web. llms.txt is a proposal that tools can adopt or ignore.
Does Google use llms.txt?
No. Google's guide to generative AI features in Search, updated on July 10, 2026, says Search doesn't use AI text files such as llms.txt, and that publishing one will "neither harm nor help your site's visibility or rankings". Its page on AI features adds that a page only has to be indexed and eligible to show with a snippet to appear in AI Overviews or AI Mode.
Does llms.txt get you cited by ChatGPT, Perplexity or Gemini?
We found no evidence that it does, from three directions.
What the AI companies document. OpenAI's crawler documentation says sites that opt out of OAI-SearchBot won't be shown in ChatGPT search answers, and it controls that through robots.txt. Perplexity's says to allow PerplexityBot in robots.txt to appear in its search results. Neither page names llms.txt as a factor, though both companies publish an llms.txt index of their own developer documentation.
A study of 300,000 domains. SE Ranking checked about 300,000 domains and found an llms.txt on 10.13% of them. It found no link between having the file and how often AI systems cited a domain, and its prediction model became more accurate when the llms.txt variable was removed.
Our own data. connu.ai has served an llms.txt since September 15, 2026. In the 30 days to October 6, 2026, Cloudflare's AI Crawl Control logged more than 2,000 requests from AI and search crawlers to connu.ai and the sites we host with it, including 517 requests for sitemap.xml. It logged 0 requests for /llms.txt.
On October 6, 2026 we also asked ChatGPT, Gemini, Google AI Mode and Perplexity 33 questions, three times each, and counted how often each of ten agencies was named in answers to questions that didn't mention it by name. Connu was named 0 times in 293 such answers. Six of the ten agency sites have an llms.txt, and the two agencies named most often don't:
| Agency | Has llms.txt | Named in answers |
|---|---|---|
| Agency A | No | 24 of 374 |
| Agency B | Yes | 19 of 374 |
| Agency C | No | 15 of 374 |
| Agency D | No | 4 of 374 |
| Agency E | Yes | 3 of 374 |
| Agency F | Yes | 1 of 386 |
| Agency G | No | 0 of 386 |
| Agency H | Yes | 0 of 386 |
| Agency I | Yes | 0 of 386 |
| Connu | Yes | 0 of 293 |
Ten sites is a small sample, and well-known agencies get named for many reasons. What the table does show is that having the file didn't decide who got named.
Who uses llms.txt
The file is useful to tools that read a site on purpose, rather than to search engines deciding what to rank or cite:
- AI assistants and agents reading documentation. The proposal's site notes that documentation platforms generate an llms.txt automatically, and OpenAI and Perplexity publish one for their developer docs.
- Chrome's Lighthouse. Lighthouse 13.2.0, released on May 1, 2026, added an Agentic Browsing category that checks whether a site's llms.txt follows the recommendations. Lighthouse describes the category as still under development and subject to change.
Should you add one?
Yes, if it takes minutes. We add one to the sites we build, generated from the same content as the pages, so it never says something the site doesn't. But don't pay anyone to optimize your llms.txt, and don't expect it to change what ChatGPT or Google's AI Overviews say about you.
What moves AI answers is the work that moves search:
- Let the crawlers in and get indexed. AI Overviews and AI Mode use pages that are indexed and eligible for a snippet, as Google's page on AI features says. ChatGPT search leaves out sites that block OAI-SearchBot, as OpenAI documents.
- Answer the questions your customers ask, with pages that rank for them.
- Be present where AI assistants look for recommendations: review sites, directories and lists in your field.
- Publish specific, first-hand facts: prices, process, timelines and measured results.
How to create one
- Write the file by hand, following the format above: the site's name, a one-sentence summary, then a short list of links to your most important pages, each with a note on what the reader will find there.
- Save it as
llms.txtin the site's root directory, so it loads atyourdomain.com/llms.txt. - Keep it in step with the site. If your site is built from code, generate the file from the same content as the pages, as ours is, so it can't drift.
How to check a site's llms.txt
- Open the site's address followed by
/llms.txtin a browser. A missing file returns an error page. - In Chrome, run Lighthouse from DevTools and look at the Agentic Browsing category.
- Run our free website score. It lists llms.txt among its AI-readiness checks, next to the checks that matter more for being found: crawler access, speed and structured data.