Free llms.txt Tool
Free llms.txt Checker & Validator
Enter a domain to fetch its /llms.txt and check the structure against the llms.txt proposal: title, summary, sections, Markdown links, and the issues that trip people up.
We fetch /llms.txt from the origin of the URL you enter and validate its structure. Public pages only.
What is llms.txt?
llms.txt is a proposed convention — described at llmstxt.org — for a Markdown file at /llms.txt that gives AI and LLM tools a short, curated map of a site’s most important content. Instead of asking a model to crawl and guess, you hand it a tidy list: the site name, a one-line summary, and links to the pages that best explain what you do.
It is optional. It is not a web standard, it is not required by search engines or AI systems, and it does not replace robots.txt or sitemap.xml. Think of it as a way to make your content easier to use if a tool chooses to read it — not as a ranking or citation lever.
What does this checker validate?
Every result is labelled PASS, INFO, WARNING, or ERROR — so a missing optional element (INFO) is never confused with an actual problem (WARNING / ERROR).
The file itself
Whether /llms.txt is reachable, the HTTP status, the Content-Type, that the body is plain text and not an HTML page, and that it is not empty or unreasonably large.
Required structure
Whether there is exactly one H1 title — the only element the proposal treats as required — and that it comes first.
Recommended structure
Whether a blockquote summary, H2 sections, and Markdown link lists are present. These are reported as INFO when absent, because a minimal file can be just an H1.
Link hygiene
Malformed Markdown links, duplicate URLs, relative vs. absolute URLs, fragment-only or javascript:/empty targets, and how many links point to .md versions of pages.
The Optional section
Whether an `## Optional` section is present. This is a convention for links an agent can skip — never a required section.
Optionally, link reachability
If you tick the box, up to 12 links are fetched from our server (with a timeout, a concurrency limit, and the same SSRF protection) to see if they resolve. A broken target link does not make the file invalid.
The overall result is one of Valid, Needs attention, No llms.txt found, or Could not check — not a 0–100 score, because the spec is still a proposal and a numeric grade would imply more precision than exists.
llms.txt format explained
The proposal defines an order for the file’s parts. Only the first one is required:
- H1 title — One `# Name` line with the site or project name. This is the only required element.
- Blockquote summary — An optional one-line `>` summary right after the H1, giving quick context.
- Body content — Optional Markdown (paragraphs, lists) with no headings, for extra context.
- H2 sections — Zero or more `## Section` headings, each grouping related links.
- Markdown link lists — Under each section, `- [name](url)` items, optionally followed by `: notes`.
- Optional section — A `## Optional` section, by convention, holds links an agent may skip for a shorter context.
A complete example (one valid shape among many — this is not the only correct layout):
# Example Company > A concise summary of what Example Company does. Additional context about the site can go here. ## Documentation - [Getting Started](https://example.com/docs/getting-started.md) - [API Reference](https://example.com/docs/api.md) ## Optional - [About](https://example.com/about.md)
A file that is only # Example Company is also valid. Sections, summary, and links are all optional additions.
Is llms.txt required for ChatGPT or AI search?
No. This is worth being precise about, because there is a lot of confident misinformation around it:
- llms.txt is a community proposal, not a mandatory web standard. Nothing requires you to have one.
- It is not a replacement for robots.txt (crawler access rules) or sitemap.xml (the list of URLs for search crawlers). Keep publishing both.
- Adding an llms.txt does not by itself get a site into an AI index, and it does not guarantee visibility, citation, ranking, or recommendation in ChatGPT or any other AI product.
- Which AI providers read llms.txt — and how — varies and changes over time. Treat any claim that a specific product 'uses llms.txt for ranking' with caution.
A reasonable way to think about it: llms.txt is a low-cost, optional courtesy to tools that choose to use it. It can make your best content easier to find and read. It is not a growth hack, and this checker will never tell you otherwise. For the signals that more directly affect whether AI systems can access your site, see the AI Crawler Robots.txt Checker and the ChatGPT Website Visibility Checker.
Common llms.txt mistakes
1No H1
The H1 (`# Site or project name`) is the one element the proposal calls required. A file without it doesn't follow the format.
2Returning HTML
Serving your normal web page (or a styled 404 page) at /llms.txt. The file is meant to be plain-text Markdown; an HTML response means there is effectively no llms.txt.
3A soft 404
A 404 page that is returned with HTTP 200. Automated clients expect either a real 404 or a valid file — a 200 that isn't a file is ambiguous.
4Broken Markdown links
`[text](url` with a missing parenthesis, a space between `]` and `(`, or empty `[]()`. These aren't parsed as links, so the target is invisible.
5An enormous file
Pasting an entire sitemap or every page on the site. llms.txt is a curated index; a content dump defeats the purpose and may be truncated by clients.
6Treating it as a robots.txt or sitemap replacement
llms.txt does not control crawler access and is not a substitute for sitemap.xml. Keep publishing both.
7Expecting a ranking or citation guarantee
Publishing llms.txt does not make an AI system crawl, index, cite, rank, or recommend a site. It is a proposal for making content easier to use, not a ranking factor.
8Putting private or sensitive information in it
/llms.txt is public. Anything you would not publish on a normal page — internal URLs, staging links, credentials, unreleased plans — does not belong in it.
How to create a good llms.txt
Prefer to fill in a form? The Free llms.txt Generator builds the file for you from these same parts, then bring the result back here to validate it.
1Start with the H1
The first line is `# Your Site or Project Name`. If you do nothing else, this alone is a minimal valid file.
2Add a one-line summary
A `>` blockquote directly under the H1: a plain sentence describing what the site is and who it is for. Recommended, not required.
3Group your key pages into H2 sections
Use `## Documentation`, `## Guides`, `## Products`, or whatever fits. Under each, a Markdown list of `- [Page name](https://full-url)` links, optionally followed by `: a short note`.
4Prefer absolute URLs and .md versions
Use full https:// URLs. Where you publish clean Markdown versions of pages, link those so a client gets content without HTML noise.
5Use an Optional section for secondary links
Put links a client can safely skip under `## Optional`.
6Keep it short, public, and current
Only your most important pages. Nothing private. Update it when the linked pages change, and serve it as text/plain at /llms.txt.
Frequently asked questions
What is llms.txt?
llms.txt is a proposed convention (documented at llmstxt.org) for a Markdown file at /llms.txt that gives AI and LLM tools a short, curated map of a site's most important content, with links to clean versions of key pages. It is meant to make a site easier for those tools to use and understand — not to rank or promote it.
Is llms.txt required?
No. It is an optional, still-evolving proposal, not a web standard and not something search engines or AI systems require. A site without an llms.txt is not broken and is not penalised. The only element the proposal itself calls required is the H1 title, and only if you choose to publish the file at all.
Does ChatGPT use llms.txt?
Support for llms.txt varies by provider and product and can change over time. You should not assume any specific AI system reads it, and you should not rely on it as your only way of exposing content. Keep maintaining robots.txt and sitemap.xml regardless.
Does llms.txt improve AI rankings?
There is no guarantee that publishing llms.txt improves visibility, citations, rankings, or recommendations in any AI system. It can make your key content easier to locate and read if a tool chooses to use the file, but that is not the same as a ranking benefit, and this checker does not measure any AI outcome.
Is llms.txt the same as robots.txt?
No. robots.txt tells crawlers which paths they may or may not fetch — it is an access-control signal. llms.txt does the opposite job: it points interested tools toward your best content. They are complementary, and llms.txt does not replace robots.txt.
Is llms.txt the same as sitemap.xml?
No. sitemap.xml is a machine-readable list of all URLs you want indexed, for search-engine crawlers. llms.txt is a short, human-curated, Markdown summary of your most important pages for AI tools. Keep publishing a sitemap; llms.txt is not a substitute.
Where should llms.txt be located?
At the root of an origin, as /llms.txt (for example https://example.com/llms.txt). The proposal also allows an llms.txt under a subpath to cover URLs beneath that path, with agents using the most specific file. This checker looks at the /llms.txt of whatever origin you enter.
What format does llms.txt use?
Plain-text Markdown, served as text/plain. The expected order is: an H1 with the site or project name, then an optional one-line blockquote summary, then optional body text (no headings), then zero or more H2 sections, each containing a Markdown list of [name](url) links with optional notes after a colon.
What should I put in llms.txt?
Your site or project name as the H1, a short summary, and links to the pages that best explain what you do — documentation, key guides, product or API references, an about page. Link to clean Markdown (.md) versions where you have them. Keep it concise and curated rather than exhaustive.
Can llms.txt contain private information?
It should not. /llms.txt is publicly accessible, so anything in it is public. Do not include internal-only URLs, staging or admin links, credentials, or information you would not publish on a normal page.