Free tool · No sign-up
llms.txt Validator
Check that your llms.txt is something an AI tool can actually use. Enter a domain or paste the file: you get every problem, why it matters and the exact fix.
Check an llms.txt file
What this llms.txt checker looks at
An llms.txt file only helps if a tool can parse it, and the format is stricter than it looks. The llmstxt.org proposal defines one required element and a few conventions; this validator checks all of them, then checks the things that break in practice.
Structure
- One H1, first.
# Namemust be the first line. It is the only required part, and the most common failure is a file that starts with something else. - A summary blockquote. A
>line directly under the title saying what the site is. It is the line a model quotes when asked "what is this?". - Link sections. Each
##heading starts a list of- [name](url): notesentries. Prose, empty sections, H3s and non-link bullets are flagged, because tools that extract the lists drop them. - The Optional section.
## Optionalmeans "skip this if you are short on context", so it belongs last.
Links
Relative URLs (they point nowhere once the file is copied into a prompt), placeholder hosts such as example.com or staging, empty link names, plain http, duplicates, and whether you link to Markdown versions of your pages, which are far cheaper for a model to read than HTML.
How it is served
When you check a domain we fetch the live file and report the status code, redirects, size, encoding
and content type. The single most common real-world fault is a site that returns its homepage as
text/html at /llms.txt: the file "exists" in a browser but contains no llms.txt at all.
How to read and fix the results
Errors stop a tool reading the file correctly; fix those first. Warnings make it less useful than it should be. Tips are optional improvements. Every finding shows the line numbers involved and a concrete fix, and the "What a parser sees" outline shows how your sections and link counts come out after parsing. Fix, paste the new version into the Paste text tab to re-check before you upload, then run the domain check again.
If you are writing the file from scratch, look at real llms.txt examples first, and keep it short: a few dozen well-described links beat a copy of your sitemap.
llms.txt is one file. Agents look for about twenty.
A valid llms.txt is a good start. An assistant deciding whether it can use your site also checks robots rules, Markdown versions of pages, an MCP server and the .well-known discovery documents.
Score the whole site
from $9/month
WebAgentScan checks every agent surface, your MCP server and its security, scores the site out of 100 and re-checks it as it changes.
See WebAgentScanInstall all of it
$39 once · WordPress $49 once
The AI Indexing Pack ships llms.txt, llms-full.txt, Markdown negotiation, an MCP server and the discovery files, with a checklist to verify each one.
See the packStarting a new site?
AIWebBuilding
Sites built with AIWebBuilding publish a generated llms.txt, Markdown pages and structured data from day one, with nothing to configure.
See AIWebBuildingFurther reading
llms.txt validator FAQ
What is llms.txt?
A plain-text Markdown file at the root of a website (/llms.txt) that gives AI tools a short, curated map of the site: a title, a one-line summary, and lists of the pages that matter with a note on what each contains. It was proposed at llmstxt.org in 2024. More on what it is and who reads it.
Does Google use llms.txt?
Google has said it does not use llms.txt for Search ranking, and no major AI company has committed to reading it on every request. It is still read by coding assistants, agent frameworks and tools that load a site into a model's context, and it costs almost nothing to publish. Treat it as cheap insurance, not a ranking lever.
What does this validator check?
The structure the format defines (one H1 first, an optional > summary, ## sections containing - [name](url): notes lists, the special ## Optional section), the quality of every link, and, when you check a live domain, how the file is served: HTTP status, content type, charset, redirects and size.
Is my file or domain stored?
No. Pasted text is checked entirely in your browser and never sent anywhere. When you enter a domain, our server fetches that one public file, returns it to your browser and keeps nothing. There is no account and nothing is published.
What is the difference between llms.txt and llms-full.txt?
llms.txt is the short index of links. llms-full.txt is an optional companion containing the full text of the important pages in one file, so a tool can load everything in a single request. You can paste either into this validator; the link-list rules apply to llms.txt.
Why does my llms.txt come back as text/html?
Usually because the file does not exist and your CMS or framework is answering the request with a normal page (often the homepage or a "not found" page) with status 200. Upload the file to the web root, next to robots.txt, and make sure no rewrite rule catches /llms.txt.