The llms.txt playbook: how to maximise AI citation

The llms.txt playbook: how to maximise AI citation

AI search engines read pages to extract meaning, not just to index them. Without guidance, they start at the homepage and follow links — which means deep pages and new pages may never be read at all. llms.txt is the file that tells AI crawlers which pages to prioritise and what they contain. Here is what to put in it and how to maintain it. Test how your pages score for AI citation with our free AIO Inspector tool — it checks speakable schema, FAQ schema, internal links, heading hierarchy and content density.

The two files

There are two files, not one.

llms.txt is the curated short list, under 100 links. Grouped by section (Services, Industries, Blog, Tools, About). Each link has a short description. This is the file AI crawlers read first.

llms-full.txt is the full content of every page listed in llms.txt, concatenated into one Markdown file. This is what an AI system reads when it wants the full picture of your site.

Think of llms.txt as the index and llms-full.txt as the library. To understand how AI systems parse and score pages for citation, read our breakdown of how Google AI Overviews and ChatGPT select sources — the scoring logic applies to llms.txt optimisation directly.

The format for llms.txt

llms.txt is plain Markdown. One H1 with the site name, a blockquote with a one-sentence summary of the site, then H2 sections for each major area, with link entries under each section.

# Acme Corp

> Acme Corp builds B2B SaaS tools for operations teams.

## Services

- [Service Name](https://example.com/services/service-name) — One sentence describing this service.
- [Service Name](https://example.com/services/service-name) — One sentence describing this service.

## Blog

- [Post Title](https://example.com/blog/post-slug) — One sentence describing this post.

The description after each link matters. AI systems use it to understand what the page is about before they read it.

What goes in llms-full.txt

For each link in llms.txt, append the full content of that page in Markdown. Strip navigation, footers and sidebars. Keep the content. The description in llms.txt tells AI what the page is; llms-full.txt gives it the full text to read.

The maintenance schedule

Refresh llms.txt monthly. Every time you publish a new high-value page, add it to llms.txt and its full content to llms-full.txt. A stale llms.txt means new pages are invisible to AI crawlers until they are discovered by following links from older pages. Complement llms.txt with speakable schema on your key pages — it marks the specific sections AI should read, not just the pages AI should visit.

Add both files to robots.txt:

Allow: /llms.txt
Allow: /llms-full.txt

Allow AI crawlers in robots.txt:

Allow: /llms-full.txt

llms.txt and sitemap.xml are complementary, not substitutes. sitemap.xml is for search engines. llms.txt is for AI crawlers. Ship both.