Free llms.txt Validator
Enter a domain and we fetch its /llms.txt, or paste a file. The same checks our generator runs on an existing file: the H1 comes first, a summary follows, every list item is a real link, no stray headings, a sane size, and a sample of the links still answer.
or paste the file
See who AI recommends in your category
A valid file is housekeeping. Hypercitations shows which brands ChatGPT, Claude, Google AI Overviews and Perplexity name for a category, which sources those answers cite, and what to change.
No credit card to sign up.
What llmstxt.org asks for
The format is short. The file starts with an H1 holding the site or project name; it is the only required element. An optional blockquote on the next non-empty line gives a one-line summary. After that come zero or more Markdown paragraphs of detail, with one rule: no headings. Then H2 sections, each a list of links written as "- [title](url): note", where the note after the colon is optional. A final H2 named Optional holds links a reader can skip when it needs a shorter context. That is the whole specification, and every check this validator runs maps to one of those sentences.
What each error and warning means
Errors break the structure: no H1 anywhere, a list item in a section that is not a "[title](url)" link, a file served as HTML, a file over 500 KB, or a sampled link that returns an error. Warnings are things a reader copes with but should not have to: an H1 that is not the first line, no blockquote summary, H3 or deeper headings or a second H1 in the details, an H2 with no items, or a file over 100 KB. The one info line notes a file with no sections at all, which is valid but gives a reader nothing to follow.
Link sampling fetches up to five same-origin links from the file and reports their status. A file that lists pages that no longer exist sends readers to 404s, which is worse than a shorter file. Pasted text is linted only; enter the domain to have links checked.
No file yet, or a file that needs rebuilding
If the domain has no llms.txt, or serves the default file Shopify provides, the generator builds one from the site’s own sitemap, titles and descriptions, with Shopify collections, products and policies mapped in. And a valid file only helps if the crawlers can fetch it: the crawler checker reads robots.txt and the firewall the way they do.
A file that passes
# Example Store
> Fragrance-free skincare for sensitive skin.
## Collections
- [Cleansers](https://example.com/collections/cleansers): Low-foam cleansers.
- [Serums](https://example.com/collections/serums)
## Optional
- [Blog](https://example.com/blogs/news): 12 articlesQuestions
The validator says my file is HTML. I can see the text in my browser.
The URL /llms.txt is returning a web page, usually the site’s 404 page or homepage with a 200 status, which is a soft 404. The browser renders it, but a crawler asking for a Markdown file gets markup. Serve the file as text/plain or text/markdown at the root.
Is a file with warnings still used?
Yes. Warnings mark things a reader copes with. Errors mark structure that a reader cannot parse, like a list item without a link. Fix errors first; warnings are polish.
Why does it say the links are not fetched?
Pasted text has no domain to fetch from, so only the format is checked. Enter the domain instead and the validator fetches the live file and samples up to five of its links.
Does a valid llms.txt get me into AI answers?
No AI search vendor documents reading the file yet, and Google says it does not use it. A valid file is cheap to keep right and useful to any tool or agent that does read it. Being cited depends on crawlers reaching your pages and on what the pages say.