llms.txt is a markdown file at the root of your site that tells an AI engine what the site is and which pages are worth reading. It is a curation file: robots.txt says what a crawler may fetch, llms.txt says what is worth fetching. The two are complementary and neither replaces the other.

Where do I put the file?

At the very top level of your domain, reachable at https://yourdomain.com/llms.txt — the same place robots.txt lives. A file in a subfolder will not be found, because engines look for it at a fixed path.

Will adding llms.txt make ChatGPT cite me?

No, and be sceptical of any tool that promises that. llms.txt is a proposed convention rather than a ratified standard, and support across engines is uneven and changing. What it reliably does is remove ambiguity about what your site is and which pages matter, which is worth doing whether or not a given engine reads the file today.

Does this tool send my content anywhere?

No. Every toolbox tool runs entirely in your browser — the text you type never leaves your device, and there is no account, no upload and no server call. You can confirm it by opening this page with your network tab open.

How many links should I include?

Fewer than you think. The file works because it is a shortlist, so include the pages that genuinely explain what you do — docs, pricing, methodology, key guides — and put the rest under the Optional heading. A list of every URL on the site is a sitemap, and you already have one of those.

← All tools

llms.txt Generator

An llms.txt file is a short markdown file at the root of your site that tells an AI engine what you are and which pages to read first. This builds one in your browser — nothing you type is uploaded.

How to use this

1Enter your site or product name — it becomes the title, the one part the format requires.
2Write a one-line summary. This is the sentence an AI engine is most likely to quote back.
3Add the handful of pages that genuinely explain what you do, one per line.
4Download the file and upload it to the top level of your site, so it answers at /llms.txt.

Site or product name (becomes the title — required)One-line summary (the line an AI engine is most likely to quote)Anything else worth stating plainly (optional)

Section 1 headingLinks — one per line, as Title | URL | note

Optional section — secondary pages an engine may skip

3 things to fix

· A site name is required — it becomes the H1, the one element the format cannot omit.
· No summary. The blockquote under the title is the line an engine is most likely to quote back.
· No links yet. A file with no links tells an engine what you are called but not what to read.

llms.txt

# Your site name

What the file is actually for

An AI engine reading your site for the first time has a hard problem: your homepage is built for people, so the thing that matters most is often wrapped in navigation, banners and marketing copy. llms.txt is a plain-markdown answer to that — a title, a one-line statement of what the site is, and a curated list of URLs worth reading, with no markup to strip.

It is deliberately not a sitemap. A sitemap is exhaustive and machine-generated, and it answers “what exists here”. llms.txt is short and hand-curated, and it answers “what should I read first”. Sites that paste their whole URL list into it get the worst of both: too long to be a shortlist, less complete than the sitemap they already publish.

The honest status of this convention

llms.txt was proposed in 2024 by Jeremy Howard and is set out at llmstxt.org. It has been adopted by a growing number of sites, but it is not a ratified standard and engine support is uneven. Publishing one does not guarantee that any particular assistant reads it, and no tool can honestly tell you otherwise.

It is still worth doing, for a reason that does not depend on adoption: writing the file forces you to state, in one line, what your site is and which five pages matter. Most sites cannot do that on demand, and the ones that can tend to be clearer everywhere else too. The cost is ten minutes and one file.

Common mistakes

Relative links. The file is read away from the page it describes, so /pricing resolves against nothing. Use absolute URLs.
Putting it in a subfolder. Engines look at a fixed path. It has to answer at /llms.txt.
Treating it as robots.txt. It grants no permissions and blocks nothing. If AI crawlers are disallowed in robots.txt, this file changes nothing — check that first with the robots.txt generator.
A summary nobody would quote. “We are a leading provider of innovative solutions” tells an engine nothing it can repeat. Say what the thing is and who it is for.

Questions

What is llms.txt?: llms.txt is a markdown file at the root of your site that tells an AI engine what the site is and which pages are worth reading. It is a curation file: robots.txt says what a crawler may fetch, llms.txt says what is worth fetching. The two are complementary and neither replaces the other.
Where do I put the file?: At the very top level of your domain, reachable at https://yourdomain.com/llms.txt — the same place robots.txt lives. A file in a subfolder will not be found, because engines look for it at a fixed path.
Will adding llms.txt make ChatGPT cite me?: No, and be sceptical of any tool that promises that. llms.txt is a proposed convention rather than a ratified standard, and support across engines is uneven and changing. What it reliably does is remove ambiguity about what your site is and which pages matter, which is worth doing whether or not a given engine reads the file today.
Does this tool send my content anywhere?: No. Every toolbox tool runs entirely in your browser — the text you type never leaves your device, and there is no account, no upload and no server call. You can confirm it by opening this page with your network tab open.
How many links should I include?: Fewer than you think. The file works because it is a shortlist, so include the pages that genuinely explain what you do — docs, pricing, methodology, key guides — and put the rest under the Optional heading. A list of every URL on the site is a sitemap, and you already have one of those.

For the longer version — how engines actually use the file, and what changes when they do — read the llms.txt guide. To check whether AI crawlers can reach your pages at all, run the LLM visibility checker.

Fixing one page? Audit the whole site.

These tools work on a single page, in your browser. The full live-URL audit — SEO, GEO and entity authority — runs free with no sign-up and hands you your top fix; sign in with Google (also free) for the rest of the fixes and your saved report history.