llms.txt Generator

Free llms.txt generator for any website.

Enter your domain. We read your sitemap, pick the pages that matter, and write a spec-compliant llms.txt you can publish today. You get the whole file, not a preview.

No login. No email. 3 free files a day.

The format

What is llms.txt?

It is a single markdown file at the root of your domain. A title, a one-line summary, then your key pages grouped into sections with a short note on each. When a model needs to know what you publish, it reads that instead of guessing from your navigation.

It is markdown, not config

No rules or directives. Just headings and links a model can read in one pass.

It is a shortlist, not an archive

The point is to name your important pages. A dump of every URL is a sitemap, and models already have one.

It is a proposal, not a standard

Adoption is still early and no search engine is required to use it. The file costs minutes to publish and nothing to maintain.

Three files, three jobs

How llms.txt differs from robots.txt and sitemap.xml

These do not replace each other. Most sites should have all three.

robots.txtsitemap.xmlllms.txt
Written forCrawlersSearch enginesLanguage models
SaysWhat you may crawlEvery URL you haveWhat your key pages are about
FormatDirectivesXMLMarkdown
Includes descriptionsNoNoYes
RequiredNoNoNo
How the generator works

Three steps, about ten seconds.

  1. 1

    We find your pages

    We read robots.txt and your sitemap, including nested sitemap index files. Sites without a sitemap fall back to the links on your homepage.

  2. 2

    We pick the ones that matter

    Login, cart, dashboard, tag and pagination pages are dropped. What remains is ranked and balanced so every part of your site is represented.

  3. 3

    We write and check the file

    Titles are cleaned of brand suffixes, missing descriptions are written, pages are grouped into named sections, and the result runs through our llms.txt validator.

Output quality

Most generators dump your sitemap and call it done.

That is why their output lists your login page under a heading called Kb. Ours is built to be read.

Sections with real names

Pages are grouped as Documentation, Pricing and Customers, not as raw URL fragments.

No login or checkout pages

Transactional and access-controlled pages are filtered out before the file is written.

Descriptions that say something

Empty and placeholder meta descriptions are rewritten instead of copied through.

No watermark in your file

The file is yours. We do not add a link back to ourselves inside it.

A validator, not just a generator

Every file is checked against the llmstxt.org spec and scored, with anything that would fail listed out.

The whole file, free

No signup wall, no truncated preview, no email required to see the rest.

Publishing the file

How to add llms.txt to your website

It must be served at yourdomain.com/llms.txt. Same place as robots.txt.

Next.js, Nuxt, Astropublic/llms.txt
WordPressSite root, not wp-content
ShopifyUpload in Content, then redirect /llms.txt to it
WebflowAdd under Site settings, custom code or file upload
cPanel or shared hostingpublic_html/llms.txt
Static hostingCommit it to your public folder

Once it is live, open yourdomain.com/llms.txt and confirm you see plain text.

Questions

No. It is a proposal published at llmstxt.org in 2024. It has not been ratified by the W3C or IETF, and no AI provider is obliged to read it. Adoption is growing but still early.

On its own, no. An llms.txt makes your site easier to understand once a model reaches it. Getting reached depends on whether your pages are crawlable, cited elsewhere, and worth quoting. Treat this file as one small piece of that.

No. Google has said it does not use llms.txt for search. It has no effect on your traditional rankings, positive or negative.

Enter your domain above and we build one for you. To write it by hand, start with an H1 for your site name, add a blockquote summary, then group your key pages under H2 headings as markdown links with a short note on each.

Up to 150, chosen by importance and balanced across your sections. That is a deliberate limit. llms.txt is meant to be a shortlist of what matters, and a file listing thousands of URLs is less useful to a model, not more.

We still read your whole sitemap, including nested sitemap index files, then include the 150 most important pages. You will see how many we found and how many made the file. Large documentation sites often split into several llms.txt files by product, which the spec allows.

Yes. Every file runs through our validator before you see it. It checks the H1, the blockquote summary, the section structure, absolute URLs, duplicates and missing descriptions, then scores the file out of 100 and lists anything worth fixing.

Yes. The generator works from any public URL regardless of platform. Where the file goes differs by platform, so see the table above for WordPress, Shopify, Webflow and Next.js.

A companion file holding the full text of your pages rather than links to them. It is much larger and needs regenerating whenever content changes. Start with llms.txt.

After you add or restructure significant content. For most sites that is every month or two. The file goes stale quietly, which is the usual reason it stops being useful.

We fetch your public pages to read their titles and descriptions, then build the file. We do not require an account and we do not publish anything about your site.

Publishing the file is the easy part.

Knowing whether it changed anything is harder. Writesonic tracks the prompts your buyers actually ask across ChatGPT, Perplexity, Google AI Overviews and five more engines, shows how you compare with competitors in those answers, and tells you what to fix next.