Free tool
llms.txt checker
Quick answer
Paste your llms.txt or enter your website, and this checker validates the file against the llmstxt.org format: an H1 first, a blockquote summary, no headings in the details, H2 sections made of [name](url) links, and the Optional section last. Every problem is listed with its line number and how to fix it.
What counts as an error
Errors break the format: no H1 at the top, or more than one H1. Fix these first; a tool that parses llms.txt may read nothing else.
What counts as a warning
- No blockquote summary after the H1.
- H3 or deeper headings, which the format does not use.
- List items in a section that are not [name](url) links.
- Relative links, which point nowhere once the file is read on its own.
- The same URL listed twice, or a section with no links.
- An Optional section that is not the last one.
The rules are the format published at llmstxt.org. Checking a live site fetches /llms.txt from the root of the address you enter; pasting checks the text in your browser and sends nothing.
No file yet? The llms.txt generator builds one that passes.
Frequently asked questions
What makes an llms.txt file valid?
It must start with an H1 naming the site or project. After that, in order: an optional blockquote summary, optional details with no headings, and optional H2 sections whose items are links written as [name](url), each optionally followed by a colon and a note.
Can llms.txt use H3 headings?
The format only defines an H1 and H2 sections, and says the details section contains no headings. A model may still read H3s, but tools that parse the format will not expect them.
Should links in llms.txt be absolute?
Yes, use full URLs. The file is often read on its own, away from your site, where a relative link points nowhere.
Why does the checker say my llms.txt was not found?
It requests /llms.txt at the root of the address you entered. A 404, a redirect to an HTML page or a file served behind a login all count as not found.