llms.txt, explained in five minutes

Every website has a small file called robots.txt. It tells search engines which pages they may read. It has been there since 1994 and nobody thinks about it.

llms.txt is the same idea for AI assistants. It is a plain-text file at the root of the site, at yourdomain.com/llms.txt, that says in ordinary language what the business is, what it offers and how to reach it.

Why a summary beats a homepage

A homepage is built for a person with eyes: a hero image, a headline, a menu, a few cards. An assistant has to reconstruct the business from that, and it gets things wrong. Prices get missed. A city in the footer gets read as the only city served. A tagline gets taken literally.

llms.txt removes the guessing. The business writes its own summary in the order that matters: who, what, where, how much, how to contact. It gives any system that reads it the facts in the business's own words.

What goes in

Ours is short, and the shape is the same for every client:

• One line naming the business and what it is.
• One line on where it works.
• The offers, with prices where prices are public.
• How the work happens, in a sentence or two.
• Contact details.
• A list of the articles on the site, so the assistant can cite them.

That is it. A few hundred words. No marketing language, because the reader is a machine that ignores adjectives and quotes facts.

What to leave out

Anything that changes weekly. Anything you would not want quoted verbatim. Anything that is not true yet. An assistant that reads "24-hour emergency service" will tell a customer at 2 a.m. that you answer the phone.

How ours stays accurate

The trap with a file like this is drift: the site changes, the file does not, and the assistant quotes last year's prices. We update llms.txt in the same change as the page, so the prices agree. The article list is appended by the same pipeline that publishes the article.

You can read this site's own file at aurumdesigns.co/llms.txt. It is the version of this business that an assistant sees.

Does it work?

Support is uneven and changing fast. Google says Search does not need it, and no major assistant documents reading it. It costs almost nothing and does no harm, so we ship it, but we sell it as a tidy summary, not a way into AI answers.

The free AI visibility check looks for an llms.txt file without scoring it, alongside the signals that do count: crawler access, structured data, plain-text facts and service pages.

FAQ

What is llms.txt?

llms.txt is a plain-text file at the root of a website, at yourdomain.com/llms.txt, that says in ordinary language what the business is, what it offers and how to reach it. The business writes the summary itself, in the order that matters: who, what, where, how much and how to contact.

Does Google use llms.txt?

Google says Search does not need it, and no major assistant documents reading it. Support is uneven and changing fast. We include one because it costs almost nothing and does no harm, but we treat it as a tidy summary, not a way into AI answers.

Do I need an llms.txt file?

It is optional. A site does not need one to appear in Google Search. Every site Aurum Designs builds ships with one, and we update it in the same change as the page so the prices agree.

What should be included in an llms.txt file?

One line naming the business and what it is, one line on where it works, the offers with prices where prices are public, a sentence or two on how the work happens, contact details, and a list of the articles on the site. A few hundred words with no marketing language. Leave out anything that changes weekly, anything you would not want quoted word for word, and anything that is not true yet.

What is the difference between llms.txt and robots.txt?

robots.txt tells search engines which pages they may read, and it has been in use since 1994. llms.txt is a plain-text summary that says what the business is, what it offers and how to reach it.