# llms.txt generator and validator (free) · EarnCitations

> Draft an llms.txt from your own pages in seconds, then validate its structure against the llmstxt.org format. Free, no signup.

Source: https://earncitations.com/llms-txt.html

> Every page on this site: https://earncitations.com/llms.txt (the whole site as one file: https://earncitations.com/llms-full.txt)

Free tool updated 2026-11-22

## llms.txt generator and validator

Draft an /llms.txt from your own pages, then check its structure. In our scan of 117 software sites, 76 published a valid llms.txt and 24 an llms-full.txt ([research](https://earncitations.com/research.html)).

### Generate from your site

We read your homepage and key pages (as [EarnCitationsAudit](https://earncitations.com/bot.html), honouring robots.txt) and draft the file from your own titles and descriptions. Anything we can't learn from your pages is marked TODO.

```

```

Review it, replace every TODO, and publish it at /llms.txt. The draft is also loaded into the validator below.

### Validate

Paste any llms.txt. Checks run in your browser; nothing is sent.

### The format in one minute

```
# Your name                       ← required: one H1, first line
> One sentence: what you do, for whom.   ← recommended summary

Optional paragraphs with key facts.

## Section name                    ← sections of links
- [Page title](https://…): what's there
## Optional                        ← links tools may skip
- [Blog](https://…)
```

### Questions

#### What is llms.txt?

A Markdown file at /llms.txt that summarises a site for language models: an H1 with the name, a one-line summary in a blockquote, then sections of links to the pages that matter, each with a short note. It is a proposal published at llmstxt.org, not a formal standard. Version 2 of the proposal (modified 2026-08-10) also lets a file sit at a path such as /docs/llms.txt, covering the pages under it, and allows a byte-order mark before the H1.

#### Do AI search engines use llms.txt?

Support varies and changes. The proposal's author writes that it has mostly been used by agents at the time they answer, coding agents following documentation above all, rather than for training. It is cheap to publish, and it forces you to decide which pages define you. Treat it as one signal among many, not a ranking trick.

#### What is llms-full.txt?

A convention some sites follow, not part of the llms.txt proposal: the full text of your key pages in one file, so a tool can read everything in one request.

#### Can I offer Markdown copies of my pages?

Version 2 of the proposal suggests a Markdown copy at the page's URL with .md added (or replacing the extension), pointed to from the page with <link rel="alternate" type="text/markdown"> or the same as an HTTP Link header. The audit (for your pricing page) and the [crawler view](https://earncitations.com/view.html) (for any page) follow either, and check the copy has every price the page shows people; the [Delivery check](https://earncitations.com/delivery.html) checks your new text reached the copy too. Build the copies in your deploy so they never fall behind: node bin/earncitations.mjs md-copies dist --site https://you.com --link writes one next to every built page and adds the link tag. We do the same for this site.

#### Should llms.txt list every page?

No. List the pages that answer buyer questions: pricing, product, docs, comparisons, about, contact. Put secondary links under "## Optional".

llms.txt is one of about 40 checks in the [free AI-readiness audit](https://earncitations.com/#audit).
