Free tool · No email needed

Free llms.txt generator

Build an llms.txt from what you already publish: your site’s name and summary from your homepage, and up to 20 pages from your own sitemap. Every link is a page you already publish; nothing is invented.

We read your homepage and your sitemap: robots.txt, then /sitemap.xml and /sitemap_index.xml. The pages it lists are not fetched. A large sitemap can take a few seconds.

Live analysisReady
–pages listedUp to 20, chosen from your sitemap
  • Not checked yet: Your site’s name, from your homepage’s own tags
  • Not checked yet: A one-line summary, from your meta description
  • Not checked yet: Pages chosen from your own sitemap
  • Not checked yet: Only links on your own domain
  • Not checked yet: Nothing written by a model, nothing invented
How it works

Get your llms.txt in three steps

  1. Step 01

    Enter your site

    Give us example.com. There is no account to make and no email to hand over.

  2. Step 02

    We read your homepage and sitemap

    Your name and summary come from your homepage’s own tags, the pages from your own sitemap. The pages listed are not fetched.

  3. Step 03

    Copy or download the file

    Read it, edit what you’d say differently, and save it as llms.txt at the root of your site.

Everything you need to know about llms.txt

Four short chapters on what the file is, what it is not, who reads it, and how this generator decides what goes in yours.

Chapter 1: What llms.txt is

A proposed Markdown file that points agents to your best pages.

llms.txt is a proposal by Jeremy Howard (opens in a new tab), first published in September 2024 and revised as v2 in August 2026, to give AI agents a short, clean guide to a website. Web pages are built for people and wrap their information in navigation and scripts; the file is a plain Markdown overview an agent can read first and follow links from.

The format is small. The proposal asks for a top-level heading with the name of the project or site, which is the only required part, then a blockquote with a short summary, optional paragraphs, and sections under second-level headings that each hold a list of links, written as a Markdown link with an optional note after a colon. A file can sit at the site root or, under v2, in a subfolder, where it covers the pages under that path. The lesson on what llms.txt is and what it isn’t walks through one for a small business.

Chapter 2: What it is not

Not a ranking factor, and not a permission.

It is not something Google Search uses. Google’s guide to generative AI features (opens in a new tab) says you don’t need to create new machine-readable files or Markdown to appear in Google Search, its generative AI features included, because Google Search doesn’t use them. It adds that making an llms.txt for other services is completely fine, and that doing so will neither harm nor help your visibility or rankings in Google Search.

It is also not access control. The proposal itself draws the line: robots.txt tells automated tools what access is acceptable, while llms.txt is read on demand, when an agent wants information. A crawler you block in robots.txt stays blocked whatever this file lists, so write your robots.txt rules for AI crawlers first, and test them with the robots.txt checker. Our longer piece on whether you need an llms.txt at all weighs the hour it costs against the same hour spent elsewhere.

Chapter 3: Who reads it

Each service decides for itself.

No service is obliged to read the file, and publishing one doesn’t guarantee that any assistant will fetch it, prefer it or cite you. The proposal reports (opens in a new tab) that thousands of sites publish one and that the files are used most heavily for software documentation, where coding agents follow them to API references and tutorials. That is real use, in one kind of site; it is not evidence about yours. Check the documentation of the assistant you care about before counting on it.

The v2 revision (opens in a new tab) also states how a file is meant to be used: an agent views or searches it, then follows the relevant links, which should point to LLM-friendly content such as Markdown versions of pages. This generator lists the pages your sitemap has, which are your ordinary web pages. If you publish Markdown versions, swap those links in.

Chapter 4: How this generator builds yours

Your tags, your sitemap, and no model.

The name is your homepage’s og:site_name, or else a name read from its <title>, or else your domain. The summary is your meta description, or else your title when it says more than the name. When your homepage declares neither, the summary is a placeholder for you to replace, and the result says so. The lesson on saying what you are in one sentence is the place to start if that line is hard to write.

The pages are URLs from your own sitemap, on your domain or a subdomain of it, with query strings dropped and each page listed once: the root first, then the shallowest paths, up to 20. Labels are made from each URL’s last path segment, not from the page. We don’t fetch the pages listed, and no AI model writes any part of the file, so it can’t describe a page that isn’t there. It can still list one you’d rather leave out, which is why the file is a draft for you to read, not a finished job.

Best practices

What a useful llms.txt gets right

  • Do: Read it before you publishThe file is built from your tags and your sitemap, not from an understanding of your business. Check every line says what you would say.
  • Do: Write the summary yourselfOne sentence on what you are and who you serve. If the generator left a placeholder, replace it; if it borrowed your meta description, see that it still fits.
  • Do: Keep the list to pages that answer questionsRemove login, cart, legal and thank-you pages, and add the pricing or docs page your sitemap may list too deep to be picked.
  • Do: Serve it as plain text at the rootPut it at example.com/llms.txt and open that address in a browser to see the file load as text.
  • Do: Give it an owner and a review dateA list of pages goes out of date when pages move. Decide who re-reads it, and when.
  • Do: Settle crawler access firstThe file suggests pages; it opens nothing. A crawler your robots.txt or firewall blocks stays blocked.
Common mistakes

What goes wrong most

  • Avoid: Treating it as a ranking leverGoogle says Google Search ignores llms.txt, AI features included. Publish it for the services that read it, not for rankings.
  • Avoid: Expecting it to grant or block accessIt is a list of suggestions. Permission lives in robots.txt and in whatever your CDN or firewall does with bots.
  • Avoid: Leaving the placeholder inA line that reads “Add a one-line summary” tells a reader the file was never looked at.
  • Avoid: Listing pages that movedLinks that redirect or return 404 make the whole file less useful. Test them when you review it.
  • Avoid: Adding pages or claims that aren’t on your siteA link to a page you never published, or a description of a product you don’t sell, is worse than a short file.
  • Avoid: Publishing it and forgetting itA stale file names old prices and removed pages under your name. A thin current file is better.
FAQ

Questions people ask

What is llms.txt?

A proposed Markdown file, usually at /llms.txt, that points AI agents to the most useful pages of a site. The only required part is a top heading with the site’s name; a short summary and sections of links usually follow. It is a proposal published at llmstxt.org, not a standard every service follows.

Will an llms.txt help me rank in Google or appear in its AI answers?

No. Google’s guide to generative AI features says Google Search doesn’t use these files, and that creating one will neither harm nor help your visibility or rankings in Google Search. If you publish one, do it for other services that read them.

Which AI assistants read llms.txt?

That is each service’s own decision, and publishing a file doesn’t oblige any of them to read it. The proposal says the files are used most heavily for software documentation, where coding agents follow them. Check the documentation of the assistant you care about before counting on it.

Where do the links in my file come from?

From your own sitemap, found through robots.txt or at /sitemap.xml and /sitemap_index.xml. We keep URLs on your domain and its subdomains, drop query strings, and list up to 20 with the shallowest paths first. We don’t fetch those pages, so check that each one still loads.

What if my site has no sitemap, or my homepage blocks you?

You still get a valid file. With no sitemap it names your homepage alone, and we never guess at pages like /about. If your homepage can’t be read, the name comes from your domain and the summary is a placeholder for you to replace. The result tells you which of these happened.

Where does the file go once I have it?

Save it as llms.txt at the root of your site, so it loads at example.com/llms.txt as plain text. The proposal also allows a file in a subfolder, which then covers the pages under that path.

Is the llms.txt generator free?

Yes, with no sign-up and no email, and we don’t keep the result. No AI model writes any of it: the name, the summary and the links come from your own tags and sitemap. A quick verification that is usually invisible runs with each request, to keep automated abuse out.

Keep going

More free tools

All free tools

See what AI says about your brand, whatever it reads to get there