LLMs.txt Generator

What llms.txt is, and what it is not

llms.txt is a proposed convention for giving language-model tools a short, curated map of public pages. It is not a robots.txt replacement, an access-control file, or a guarantee that an assistant will read or cite the listed links.

Use the main file for a short description and a small set of canonical links. Put longer reference material in normal HTML or Markdown pages that have stable URLs. The optional llms-full.txt download requires your own full-content notes.

Before publishing

• Replace every sample URL with the canonical production URL.
• List only public pages that an unauthenticated visitor can open.
• Remove stale docs and duplicate links.
• Keep robots.txt, authentication, and server permissions as the actual access controls.

A small, complete input example

For a documentation site at https://example.com, enter a site name such as "Example Docs", a one-sentence summary of what readers can find, and two real core links. Each link uses one line in Title | URL | Description format. For example:

Getting started | https://example.com/docs/start/ | Install and configure the product
API reference | https://example.com/docs/api/ | Endpoint descriptions and parameters

These are illustration URLs, not a ready-to-publish file. Replace them with pages you own and test each link in a private window. Keep the main llms.txt short enough that a reader can scan it; put detailed explanations on the linked pages. Only create llms-full.txt when you have written original notes for that field. The tool will not export incomplete starter text as if it were your published material.

A successful download does not mean an AI assistant will discover, obey, cite, or prefer the file. Measure actual referrals and verify the accessibility of the linked resources separately.

After download, publish the file at /llms.txt, open it in a browser, and check every link. The proposal and examples are documented at llmstxt.org.