llms.txt

A simple markdown file that hands AI agents a clean map of your site instead of making them crawl your HTML. Here is what it is, how it differs from robots.txt, whether anyone reads it yet, and how to ship one.

What llms.txt is

llms.txt is a plain markdown file at the root of your site that gives a language model a curated, summarized map of your best content. Instead of crawling and parsing a sprawl of HTML, an agent can read one clean document and understand what you do and where to look.

The format is deliberately simple: a title, a one-line summary of your site, then sections of links with a short description for each. You are doing the model's first pass of work for it, and pointing it at the pages you actually want it to use. This is one piece of the broader agent legibility stack.

llms.txt vs robots.txt vs sitemap.xml

Three root files, three different jobs. You want all three.

FileAudienceJob
robots.txtCrawlersPermission: what may be accessed
sitemap.xmlSearch enginesInventory: every URL to consider
llms.txtLanguage modelsA guided tour: the best pages, summarized in markdown

A larger companion file, llms-full.txt, takes it further by concatenating the full markdown of your key pages so an agent can ingest everything in one fetch. We publish our llms.txt and our llms-full.txt, because a site that sells agent legibility without one would be a bad joke.

Does it actually work yet?

Here is the honest answer most pages will not give you. As of now, the major training and search crawlers mostly ignore llms.txt and fetch your HTML directly. There is no public commitment from the big labs to read it in production, and studies of crawler logs show requests to llms.txt are still rare. What does use clean machine versions today are coding agents like Claude Code and Cursor, which request markdown when it is offered.

So why ship it? Because it costs almost nothing, it helps the agents that do read it, it forces you to think clearly about your most important pages, and models train on the conventions that exist now. It is cheap insurance and a small head start, not a silver bullet. We would rather tell you that than sell you a myth. The higher-leverage work is structured data and entity consistency.

How to create your llms.txt

List your most important pages, grouped into a few sections, each with a one-line description in plain markdown. Lead with what you want agents to use most. Keep it curated rather than exhaustive, because the point is guidance, not a second sitemap. Our free llms.txt template and generator gives you an annotated starting point and a draft you can refine by hand.

llms.txt, answered

What is llms.txt?
llms.txt is a plain markdown file you place at the root of your site, at yoursite.com/llms.txt. It gives AI agents a curated map of your most important pages, each with a short description, so a model can understand your site without crawling and parsing all of your HTML. It was proposed by Jeremy Howard of Answer.AI and has been adopted by companies like Anthropic, Stripe, Vercel, and Cloudflare.
How is llms.txt different from robots.txt and sitemap.xml?
robots.txt tells crawlers what they may and may not access. sitemap.xml lists every URL for indexing. llms.txt does something different: it curates and summarizes your best content for a language model, in markdown a model can read directly. robots.txt is permission, sitemap is inventory, llms.txt is a guided tour.
Do AI crawlers actually read llms.txt?
Honestly, not many yet. Server logs show that major training and search crawlers still mostly fetch HTML directly, and there is no public commitment from the big AI labs to read llms.txt in production. Coding agents like Claude Code and Cursor do use clean machine versions of pages. So today llms.txt is cheap insurance and good housekeeping, not a magic ranking lever. It costs almost nothing and positions you for where this is heading.
What is llms-full.txt?
llms-full.txt is the companion file to llms.txt. Where llms.txt is a concise index, llms-full.txt is the complete content of your key pages concatenated into one markdown document, so an agent can ingest everything in a single fetch. Large documentation sites publish both.
How do I create an llms.txt file?
Start from our free annotated template, list your most important pages grouped by section with a one-line description each, and keep it in markdown. You can generate a first draft with our template tool, then curate it by hand so it points agents at what matters most.

Build your llms.txt

⋅ View this page as Markdown for AI agents