Free llms.txt Checker and Validator (v2)

What the llms.txt checker tests

First, the checker fetches the file and makes sure it's actually an llms.txt file. This catches a pretty common problem: some sites return their normal HTML page, or a "not found" page, at /llms.txt with a 200 status. Then it reads the file line by line:

Reachability: the status code, any redirects, the content type, the encoding and the file size.
The H1: the site or project name at the top. This is the only part of the spec that's required.
The summary: a blockquote under the H1 that says what the site is.
The link lists: H2 sections with list items written as - [Name](url): notes.
The links themselves: up to 50 linked URLs, each checked for its status and for a Markdown version.
Scope: which pages the file covers, based on where it sits on the site.
v2 link relations: whether your pages point to the file, and to their own Markdown versions.

How to read the result

Errors mean the file won't work as an llms.txt. Warnings mean it works, but there's a problem worth fixing. Info items are optional extras from the v2 spec.

Check	Result	Fix
File at /llms.txt	Error on a 404, a redirect loop or an HTML page	Save a plain text file at the root of the site, or at the path it should cover
H1 at the top	Error if missing	Start the file with `# Your site name`
Summary	Warning if missing	Add one line starting with `>` that says what the site is
Headings	Warning on H3 or lower, or on an H2 that isn't a link list	Use H2 only for link lists, and put any notes above the first H2 without headings
Link list format	Warning on list items without a link	Write each item as `- [Name](url): notes`
Broken links	Warning for each link that returns an error	Fix the link or remove it
Content type	Warning if it's text/html	Serve .txt files as text/plain or text/markdown
Link relations	Info	Add `<link rel="describedby" href="/llms.txt">` to your pages any time, and `rel="alternate"` links once you have Markdown versions

In general, fix the errors first, then the warnings.

What changed in llms.txt v2

llms.txt was first proposed by Jeremy Howard of Answer.AI in September 2024. The current version, v2, was last updated on 10 August 2026. The file itself barely changed: it's still Markdown with an H1, a summary, optional notes and H2 link lists. Most of v2 is about how AI agents find and use the file:

How agents use it: v2 describes agents reading or searching the file, then following its links to content written for them.
Scope: v1 allowed files in subfolders but didn't say what they covered. In v2 a file covers the pages under its path, and the most specific one wins, so a project on a shared domain can have its own file.
Markdown versions of pages: each page can have a clean Markdown copy at the same URL with .md added (page.html.md), or, new in v2, with the extension replaced (page.md).
Link relations: a page can point to its Markdown version with rel="alternate" type="text/markdown" and to its llms.txt with rel="describedby", either as an HTML link tag or in an HTTP Link header.
The Optional section: it's still allowed for secondary links, but v2 gives it no mechanical meaning. In v1 it told tools what to leave out when they needed a shorter context.
Removed: the llms_txt2ctx tool for expanding a file into one long context, which v1 described.

The checker tests the v2 rules, and reports the new link relations as optional.

An llms.txt example that passes

Here's a small file for a made-up invoicing product that passes every check:

# Acme

> Acme is invoicing software for freelance designers, with plans from $12 a month.

Acme sends invoices, tracks payments and reminds late clients by email.

## Product
- [Pricing](https://acme.io/pricing.md): Plans, limits and what each includes
- [Features](https://acme.io/features.md): What the product does
- [Docs](https://acme.io/docs/index.md): Setup and API

## Optional
- [Changelog](https://acme.io/changelog.md): Release notes by month

Keep it short. It's basically a list of your most useful pages, with one line on each. Link to the Markdown versions of your pages if you have them, and to the normal pages if you don't.

Does llms.txt do anything?

In short, it's a map for the AI agents and tools that read it. The proposal is written for agents that read a site on someone's behalf, and v2 describes them reading the file and then following its links to the right pages. There aren't any published numbers yet on how many tools actually do this, or on whether having a file changes which pages AI engines cite.

Peak Answer's crawler access study recorded what came back at /llms.txt on the tech sites ChatGPT and Google cite. A plain fetch couldn't tell a real file from other text, so Peak Answer ran this checker on the same 887 sites in November 2026: 58% had a real llms.txt, and 2% returned a web page at /llms.txt instead. The llms.txt adoption study has the full breakdown.

So think of llms.txt as a small extra. It only takes a few minutes to write, and writing it makes you spell out what you sell and which pages matter, but it won't get a page cited on its own. Crawler access matters a lot more. OpenAI says sites that block OAI-SearchBot won't be shown in ChatGPT search answers, and the AI crawler checker tests that part.

How to fix common llms.txt problems

The checker sees an HTML page. Your CMS or host is serving its normal page, or a soft 404, at /llms.txt. Upload the file as a static file at the root, or add a route that serves it as plain text.
The H1 isn't at the top. Remove anything above the title, like a comment or a blank line, and start with # Name. A byte-order mark before it is fine.
Links return 404 or redirect. Update them to the final URLs.
The file is very long. Cut it down to your most useful pages and the main sections of your docs. Secondary links can go under ## Optional.
The wrong content type. If the server sends text/html, set text/plain or text/markdown for .txt files in your server or CDN settings.
The file is out of date. Update it whenever you add or remove key pages. The llms.txt generator builds a first version from your site.

llms.txt, robots.txt and sitemap.xml

All three files sit at the root of your site, but they do different jobs:

	llms.txt	robots.txt	sitemap.xml
Job	Describes the site and points to its key pages	Tells crawlers what they may fetch	Lists the URLs you want indexed
Format	Markdown	Plain text rules	XML
Who reads it	AI agents and tools that support it	Crawlers that follow it	Search engines
Required	No	No, but every site should have one	No, but most sites should have one

You can check the other two with the robots.txt checker and the sitemap checker.

Common questions

How do I check my llms.txt file?

Enter your site above. The checker fetches /llms.txt, tests it against the v2 spec and checks the links in it. You can also open yoursite.com/llms.txt in a browser: it should show plain text that starts with `#` and your site name.

Is llms.txt still used?

The proposal is still maintained, and it was last updated in August 2026. How many AI tools actually read it isn't published.

Does llms.txt really work?

It works as a map for the tools that read it. There's no published evidence that it changes which pages AI engines cite.

Is llms.txt mandatory?

No, llms.txt is not mandatory. It's an optional proposal, and nothing breaks without it. If you do publish one, the only required part is the H1.

What should llms.txt contain?

An H1 with your site name, a one-line summary in a blockquote, a few lines of context if needed, and H2 sections listing your most useful pages as `- [Name](url): notes`. Secondary links can go under `## Optional`.

What is llms-full.txt?

It's a separate file some sites publish with the full text of their docs in one place. It isn't part of the llms.txt spec. The checker notes whether you have one, but doesn't grade it.

Track your AI visibility every week

Peak Answer runs checks like these every week, then writes the content and gets the placements.

Start 7 day trial