What does a good llms.txt look like?

One H1 with your site name, a blockquote summary with concrete specifics, sections grouped by intent (Docs, Pricing, Company), and each link carrying a real description after the colon. Curate five to ten key pages — curation is the value.

Where does llms.txt go on my server?

At the domain root so it resolves at https://yoursite.com/llms.txt. Most hosts and CDNs serve it as a plain static file — upload it next to robots.txt.

How do I know my llms.txt is being read?

The standard is young and adoption by engines is gradual. What you control: the file resolves correctly and scores green on the llms.txt pillar of the checker, and the content is quotable when a model does read it.

llms.txt Example — Annotated, Copy-Paste Ready

A complete example

# Acme Analytics

> Privacy-first web analytics for SaaS teams. Cookieless tracking, real-time dashboards, GDPR-compliant by default.

Topics covered: web analytics, dashboards, privacy, GDPR.

## Documentation

- [Quickstart](https://acme.com/docs/quickstart): Install the snippet and see live data in five minutes.
- [Dashboard guide](https://acme.com/docs/dashboard): Every metric on the main dashboard, defined.
- [API reference](https://acme.com/docs/api): Endpoints, authentication and rate limits.

## Pricing

- [Pricing](https://acme.com/pricing): Plans from $19/mo. Free for sites under 10k views.

## Company

- [About](https://acme.com/about): Founded 2021, remote team of 12, used by 4,000+ sites.
- [Blog](https://acme.com/blog): Analytics guides and product updates.

Why this works, line by line

H1 is the site name — exactly one, matching your brand. This anchors every downstream reference.
The blockquote is a real summary — what you do, for whom, with the two or three specifics (cookieless, GDPR, $19/mo) an engine can lift verbatim when describing you.
Sections group by intent — Docs, Pricing, Company. A model looking for 'what does Acme cost' lands in the right neighborhood instantly.
Every link has a description after the colon — the description is the payload. A bare URL list is barely better than no file.
Six links, not sixty — curation is the point. This is a map of what matters, not a sitemap dump.

The mistakes that waste the file

Marketing fluff in the summary ("innovative synergistic platform") — nothing liftable, nothing believed.
Links without descriptions, or descriptions that repeat the page title.
Dead or redirected URLs — worse than omitting the page.
Duplicating your whole sitemap: if everything is important, nothing is.
Forgetting the file must live at the root: https://yoursite.com/llms.txt.

Build yours in two minutes with the llms.txt generator, start from the fill-in versions in llms.txt template, and see how it differs from robots rules in llms.txt vs robots.txt. Then verify the file resolves with a free check of your domain.

llms.txt Example (Annotated)

A complete example

Why this works, line by line

The mistakes that waste the file

Frequently asked questions