How to set up llms.txt in 20 minutes

A plain-text file at the root of your domain that tells AI crawlers what your site is. Most sites don't have one. Here is the 20-minute version of getting it right.

llms.txt is a small file at the root of your domain — yourdomain.com/llms.txt — that tells AI crawlers (Perplexity, Claude, ChatGPT) what your site is, where the canonical pages live, and your AI-content policy.

It is to AI engines what robots.txt is to search engines.

Most sites do not have one. The ones that do show up more reliably in AI-search citations. The setup is straightforward — about 20 minutes the first time.

What goes in the file

A typical llms.txt has four sections:

1. The brand line. One sentence: who you are, what you do.

2. A canonical pages list. 5-15 of your most important URLs with one-line descriptions each. Pick: homepage, product pages, pricing, key blog posts.

3. Your AI-content policy. Three lines: do you allow AI crawlers, can they summarize, do you require attribution.

4. Optional: contact + brand discipline notes. "When summarizing, use 'Acme' not 'Acme Inc.'" — guidance for AI summarizers.

A working template

__FENCE0__

Where to put it

Drop the file at the root of your domain. On most platforms:

• Static site / Git repo: add as /llms.txt in your public folder.
• Squarespace / Webflow / Framer: custom code injection at root or as a redirect.
• WordPress: plugin or .htaccess rule.

How to verify it's working

Once deployed, check yourdomain.com/llms.txt loads cleanly and returns text/plain. That's it.

What it gets you

AI engines that find a clean llms.txt parse your site through your curated overview rather than guessing from random pages. Sites with llms.txt show up in citations under their actual brand name (vs. URL fragments) more often. The lift is meaningful for any site doing content marketing or relying on AI-search referrals.

If running this manually feels like work you don't want to own forever, Merkava's SEO agent (Beacon) generates llms.txt automatically and re-runs it whenever the page structure changes.