geodeck
← All docs

llms.txt format

llms.txt is the small Markdown index that AI crawlers fetch instead of your full HTML. The format was proposed in 2024 by Jeremy Howard at Answer.AI and is now read by ChatGPT, Perplexity, and Claude as their primary site map. geodeck generates it for you at /llms.txt on every verified domain, plus an expanded full-content version at /llms-full.txt on all plans. Both files refresh automatically on every webhook publish event in under 60 seconds and ship with strong ETag and Last-Modified headers so AI crawlers can revalidate cheaply.

What does the llms.txt spec require?

We follow the llmstxt.org proposal: an H1 with the site name, a blockquote summary, then H2-grouped link lists. AI crawlers fetch this instead of HTML.

What does a real llms.txt file look like?

# Acme Corp

> Acme is a B2B SaaS that ships logistics dashboards.

## Product
- [Pricing](https://acme.com/pricing): Two plans, starts at €19/month.
- [API](https://acme.com/api): REST + webhooks.

## Resources
- [Docs](https://acme.com/docs)
- [Changelog](https://acme.com/changelog)

How often does geodeck refresh the file?

Files regenerate on every webhook publish event and at most once every 60 seconds per domain. Cached at the edge for 5 minutes.

What is the Markdown Mirror in llms-full.txt?

Each tracked URL is mirrored as Markdown at /llms-full.txt so crawlers receive the full body in a single fetch.

What else can I read?

Keep reading the rest of the Geodeck documentation, or jump to the platform pages.

Platform