Skip to content

Tools

llms.txt generator — build an llms.txt for any website, free

Paste a URL. The generator reads your live pages and writes an llms.txt from what is actually on the site — business name, summary, services, and links to the pages agents look for. No signup, no AI model, nothing invented.

What the generator writes, and what it refuses to write

The file it produces has four parts: a heading with the business name, a one-line summary taken from your meta description, the site and contact address, and a linked list of key pages grouped by the sections an llms.txt is expected to cover — about, pricing, products, documentation, methodology, blog, contact, privacy, and terms.

What it will not do is fill a gap with something plausible. If there is no meta description, the summary line is omitted and "add a sentence saying what the organization does" appears in the gap list instead. The same applies to the contact address and to every section with no matching page. An llms.txt exists to tell agents true things about a site; a generator that guesses defeats the point of publishing one.

Worked example: running it on this site

Point the generator at www.visiscan.app and it returns a file headed # VisiScan, a summary line taken from the homepage meta description, the site origin, and a key-pages list linking the pages it matched — pricing, methodology, blog, privacy, and terms among them, each labelled with what an agent would find there.

The gap list is where the honest part shows. Sections with no matching page on this site come back as "no page found for" entries rather than invented links, because the generator reads a sample of pages rather than crawling every URL. That is the loop: generate, fill the named gaps by hand, publish at /llms.txt, then confirm with the checker.

VisiScan's own published file is at /llms.txt if you want to compare a finished, hand-maintained file against generated output.

Questions about llms.txt

What is an llms.txt file?
It is a plain-text file at the root of a site (/llms.txt) that gives AI agents a short, factual map of the site: what the organization does and canonical links to pricing, documentation, methodology, and policies. It is a proposed convention, not a standard any engine is obliged to read.
How much does this llms.txt generator cost?
Nothing. The generator is free, needs no signup, and does not run any AI model — it reads your public pages and assembles the file from what it finds.
How does the generator decide what goes in the file?
It fetches up to four public pages from your site, reads the business name, meta description, services, and contact address, and matches page URLs and titles against the sections an llms.txt is expected to cover — about, pricing, products, documentation, methodology, blog, contact, privacy, and terms.
Does it invent content when something is missing?
No. Anything it could not read from your site is left out of the file and listed separately as a gap for you to fill in by hand. A confident but wrong llms.txt is worse than an obviously incomplete one, so the generator never writes a placeholder that reads like a fact.
What are the limits of the generated file?
It reads a sample of pages rather than crawling the whole site, so a large site will need links added by hand. It cannot see pages blocked by robots.txt, and it does not execute JavaScript, so content injected client-side is invisible to it — the same view a crawler gets.
Will publishing llms.txt get my business cited by AI?
Not on its own. Google has said no AI-specific file is required for AI Overviews or AI Mode, and ordinary crawlability and content quality still decide most of the outcome. Treat llms.txt as an optional discovery aid for agents that choose to read it.
Where do I put the file once it is generated?
Save it as llms.txt in your site's public root so it serves from https://yourdomain.com/llms.txt as plain text, then confirm it with the llms.txt checker.

Next steps

Once the file is published, check that it parses and has real structure with the llms.txt checker, confirm AI crawlers can reach your pages at all with the AI crawler checker, and see whether the identity markup on your pages is complete with the schema markup checker. For the whole picture — whether AI engines actually name your business — run the AI visibility scanner. Background reading: how to get your business cited by ChatGPT.