llms.txt is a Markdown file at the root of a website (/llms.txt) that gives large language models a short, curated map of the site: its name, a one-line summary, and lists of the most useful URLs with brief notes. It was proposed by Jeremy Howard in September 2024 and is defined at llmstxt.org.
Start from a blank template and add your own sections and links.
You can edit this text directly. Changing the form above regenerates it.
This tool needs JavaScript. Enable it to generate and validate llms.txt files.
How to create an llms.txt file
- Scan your website. Enter your homepage URL and select “Scan website”. The tool reads robots.txt, your XML sitemap and the title and meta description of your key pages.
- Curate the sections. Rename sections, untick pages that don’t help an AI understand your site, and rewrite link notes so each one says what the page answers.
- Fix validation issues. Check the live validator below the output. Fix errors first, then warnings, until the grade is A.
- Upload llms.txt to your site root. Download the file and upload it to your website’s root folder (on WordPress, the folder containing wp-config.php) so it loads at https://yourdomain.com/llms.txt.
- Validate the live file. Open the “Validate” tab, enter your domain and confirm the published file is served as plain text and scores an A.
What does an llms.txt file look like?
A valid file has one H1, a blockquote summary, optional notes without headings, and H2 sections of links. This example follows the llmstxt.org format:
# Example Docs
> Example is an open-source library for building fast web APIs in Python.
Example requires Python 3.10+. All code samples use async handlers.
## Docs
- [Quick start](https://example.com/docs/quickstart.md): Install and run a first API in five minutes
- [Routing](https://example.com/docs/routing.md): Path parameters, middleware and error handling
## Examples
- [Todo API](https://example.com/examples/todo.md): Complete CRUD app with tests
## Optional
- [Changelog](https://example.com/changelog.md): Release history
Is llms.txt worth it in 2026? The data
| Finding | Source |
|---|---|
| Google Search does not need llms.txt: its AI optimization guide says you don’t need AI text files or Markdown to appear in Search or its AI features. | Google Search Central, updated July 2026 |
| Chrome Lighthouse 13.3 (May 2026) added an llms.txt audit in its Agentic Browsing category. Files fail without an H1 or Markdown links. | DebugBear; Search Engine Journal, July 2026 |
| About 36,000 sites published llms.txt by May 2026, 8.8× more than a year earlier. | Originality.ai data, reported by PPC Land |
| 97% of llms.txt files across 137,000 domains received no requests in May 2026; AI retrieval bots made about 1% of the requests that did occur. | Ahrefs server-log study, reported by PPC Land |
Our take: publish an llms.txt if you have documentation, tools or reference content that AI agents and coding assistants might read. It takes minutes, Lighthouse now checks it, and a clean file costs nothing. Don’t expect it to change your Google rankings or AI Overview citations; those still depend on original, well-structured pages.
What does the llms.txt validator check?
| Rule | Severity | Why it matters |
|---|---|---|
| Starts with one H1 | Error | The only required element. It names the site or project. |
| Blockquote summary after the H1 | Warning | One or two sentences an AI can quote to describe you. |
| No headings before the first H2 | Error | Free-text notes may use paragraphs and lists only. |
| Sections contain only link lists | Error / warning | Each item is “- [name](url): notes”. |
| Contains Markdown links | Error | Chrome Lighthouse fails llms.txt files with no [name](url) links. |
| Parser-safe link items | Warning | Use “-” bullets, no square brackets in link names, and unique section titles. Valid Markdown that breaks the reference llms_txt parser is flagged. |
| Absolute http(s) URLs, no duplicates | Warning | Agents may fetch links out of context, so relative URLs can break. |
| One “Optional” section, placed last | Tip | Links agents can skip when their context window is short. |
| File stays concise (under 100 KB) | Warning | Put full page text in llms-full.txt instead. |
llms.txt vs robots.txt vs sitemap.xml
| llms.txt | robots.txt | sitemap.xml | |
|---|---|---|---|
| Purpose | Curated guide to your best content | Allow or block crawlers | List every indexable URL |
| Format | Markdown | Plain-text directives | XML |
| Main readers | LLMs, AI agents, coding assistants | All crawlers | Search engines |
| Location | /llms.txt | /robots.txt | /sitemap.xml (or listed in robots.txt) |
| Controls access? | No | Yes | No |
Frequently asked questions
What is an llms.txt file?
An llms.txt file is a Markdown file placed at a website’s root that tells large language models what the site is and which pages matter most. It contains an H1 title, a blockquote summary, optional notes, and H2 sections listing links with short descriptions.
Where do I put llms.txt?
Upload llms.txt to the root of your domain so it is available at https://yourdomain.com/llms.txt. The spec also allows files in subpaths, such as /docs/llms.txt, which then cover the URLs under that path.
How do I add llms.txt to WordPress?
Generate the file with this tool, then upload it with your host’s file manager, SFTP or FTP to the WordPress root folder, which is the folder that contains wp-config.php. Some SEO plugins, including Yoast SEO and AIOSEO, can also generate llms.txt automatically; use only one method.
What is the difference between llms.txt and llms-full.txt?
llms.txt is a short index of links. llms-full.txt is a widely used companion convention that contains the full text of those pages in one Markdown file, so an AI tool can load everything in a single request. llms-full.txt is not part of the original proposal.
Does llms.txt improve Google rankings or AI Overviews?
No. Google’s AI optimization guide says you don’t need AI text files or Markdown to appear in Google Search, including its AI features, so llms.txt neither helps nor hurts rankings or AI Overviews. Its value is for AI agents, coding assistants and developer tools that fetch it on demand. It is cheap to publish, but it is not a substitute for clear, original pages.
Does Google Lighthouse check llms.txt?
Yes. Since Lighthouse 13.3 (May 2026), the Agentic Browsing category includes an llms.txt audit. It flags a missing H1, a file that is too short or has no Markdown links, and server errors when fetching the file. A 404 counts as “not applicable” because the file is optional. This validator checks all of these.
Does llms.txt replace robots.txt?
No. robots.txt controls which crawlers may access your site; llms.txt only suggests what to read. To allow or block AI crawlers such as GPTBot, ClaudeBot or PerplexityBot, use User-agent rules in robots.txt.
What goes in the Optional section?
Secondary links an AI can skip when it has limited context, such as changelogs, legal pages or archives. By convention the section is titled exactly “Optional” and placed last.
How do you know the generated llms.txt is valid?
The generator and validator follow the llmstxt.org format and are tested against the reference llms_txt parser published by the spec authors: hundreds of generated files must parse back exactly, and any file that parser rejects must be flagged by the validator. Valid Markdown that real parsers mishandle, such as “*” bullets or brackets in link names, is reported too.
Is this llms.txt generator free and private?
Yes. It is free with no sign-up. Generation and validation run in your browser. When you scan a site, our server fetches only that site’s public robots.txt, sitemap and pages, keeps the result in a 30-minute cache, and stores nothing else.
Last reviewed against the llms.txt specification.