Create an llms.txt file — a short Markdown guide that tells AI assistants what your site is and which pages matter. Enter your site's name and a one-line summary, add sections of links with optional notes, and mark less important links as Optional. The output follows the llmstxt.org format exactly and is ready to upload to your site's root. It runs in your browser.
| Part | Markdown | Required |
|---|---|---|
| Title | # Name | Yes |
| Summary | > One or two sentences | Recommended |
| Details | Plain paragraphs | No |
| Sections | ## Section then - [title](url): note | Recommended |
| Optional | ## Optional | No |
A Markdown file at /llms.txt that summarizes a site for AI models and links to its most useful pages. It's a community proposal from llmstxt.org, similar in spirit to robots.txt but descriptive rather than restrictive.
Only the H1 title. A summary and link sections are strongly recommended.
Links in a section titled Optional can be dropped when a model needs a shorter context. Put secondary material there.
No. It describes your site; it doesn't grant or deny access. Use robots.txt for that.
No. The file is generated in your browser.
llms.txt is a proposal from llmstxt.org for a Markdown file at a site's root that gives language models a concise, curated map of the site, since full HTML pages are long, noisy and often larger than a model's context. The format is fixed in its order: an H1 with the site or project name, which is the only required element; a blockquote summary; optional paragraphs of detail without headings; then H2 sections, each a list of links written as - [title](url): optional note. A section named Optional has a defined meaning — its links can be skipped when a shorter context is needed. The generator assembles exactly that structure, collapses line breaks inside single-line elements so they can't break the Markdown, strips square brackets from link titles, and omits empty sections and incomplete links rather than writing broken entries. Linking to Markdown versions of pages, conventionally the page URL with .md appended, is recommended because models parse them more reliably than HTML. The file is built in the page.