llms.txt is a plain text file at yoursite.com/llms.txt that gives AI tools a short guide to your site. It says what the site is and links to the pages that explain it best. This page has a free llms.txt generator and an llms.txt checker: jump to the generator or the checker.
Both tools run in your browser, and nothing you type is uploaded. Some people call the checker a validator. The file is written in Markdown, a simple way to mark headings and links with plain characters. Read on for the format, what Google says, and how it works on Shopify and WordPress.
The short version: the file is optional and quick to write, and Google says Google Search ignores it (Google Search Central, checked 7 Oct 2026). So first make sure AI bots can read your pages at all.
Skip to the free tools on this page.
What is llms.txt, and who proposed it?
llms.txt is a proposal for one small file that helps AI tools read a website. Jeremy Howard published it on 3 September 2024. The current text is version 2, last changed on 10 August 2026 (llmstxt.org, checked 7 Oct 2026).
The proposal starts from a simple problem. Web pages are built for people, with menus, ads and scripts around the words. An AI model can also read only a limited amount of text at once.
An AI agent is software that reads pages and takes steps for a user. The idea is that an agent opens the file and follows the link it needs. The proposal says the heaviest use is software documentation, where AI coding tools follow the file to reference pages and tutorials.
Some people type "llm txt" or "LLMs.txt". The file name is always llms.txt.
What does the llms.txt format look like?
An llms.txt file has up to four parts in a fixed order, and only the first one is required (llmstxt.org, checked 7 Oct 2026). The fifth point below is a naming convention for one link section, not another part.
- A title. One H1 heading with the name of the site. In Markdown that is a line starting with "# ". This is the only required part.
- A summary. A blockquote, which is a line starting with "> ", with a short description of the site.
- Details. Optional paragraphs or lists that help a reader understand the rest. Headings are not allowed in this part.
- Link sections. Optional sections, each under an H2 heading, a line starting with "## ". Each holds a list of links. A link is a dash, the page title in square brackets, the URL in round brackets, then an optional colon and a note.
- A section named Optional. By convention it holds secondary links that an agent can skip when it needs a shorter read.
A short example, line by line
Here is a file for a made-up shop. The generator below starts with the same example.
- Line 1, the title: # Example Store
- Line 3, the summary: > Example Store is a made-up online shop that sells handmade cotton bedsheets and towels and ships across India.
- Line 7, a section heading: ## Products
- Line 9, a link: a dash, Bedsheets in square brackets, the full page address in round brackets, then a colon and a short note.
Does llms.txt actually work?
It depends on who you want to read it. Google says Google Search ignores the file (Google Search Central, checked 7 Oct 2026). OpenAI, Anthropic and Perplexity each publish a page about their crawlers, the bots that read web pages. None of those pages names llms.txt as a file the crawlers read (OpenAI docs, Anthropic docs and Perplexity docs, checked 7 Oct 2026). By the proposal's own account, its heaviest use is software documentation read by AI coding tools (llmstxt.org, checked 7 Oct 2026).
So treat llms.txt as a low-cost extra, not as a way to rank.
- Google Search. Google's guide to its AI search features lists llms.txt among the things you can ignore. It says keeping one will neither harm nor help your rankings, and that it is fine to keep one for other services (Google Search Central, checked 7 Oct 2026). Google added this note on 15 June 2026 (Google Search documentation updates, checked 7 Oct 2026).
- Chrome Lighthouse. Lighthouse, Chrome's page audit tool, has an llms.txt check in its audits for AI agents. It treats the file as optional: a site with no file is marked not applicable (Chrome for Developers, checked 7 Oct 2026).
- OpenAI, Anthropic and Perplexity. Each company's crawler page tells site owners to control its bots with robots.txt, a file of rules for bots. None names llms.txt as a file its crawlers read (OpenAI docs, Anthropic docs and Perplexity docs, checked 7 Oct 2026). All three do publish an llms.txt file for their own developer documentation (OpenAI, Anthropic and Perplexity, checked 7 Oct 2026).
How many sites in Indian AI shopping answers publish one?
Depra AI looked at 262 sites that appear in Indian AI shopping answers to see whether each one publishes an llms.txt file. The result is in our post on what llms.txt is and whether it works.
Do you need an llms.txt file?
No. llms.txt is a proposal, and nothing breaks if you do not have one.
It is worth the time for a site with documentation or long help pages. It also suits a store that wants one clean page stating what it sells, its prices and its policies. A Shopify store already has a file, because Shopify serves /llms.txt for every store by default (Shopify Help Center, checked 7 Oct 2026).
It can wait if AI bots cannot read your site at all. That is decided by robots.txt. The AI crawlers list names each AI bot and gives the robots.txt lines to use.
llms.txt vs robots.txt vs sitemap.xml: what is the difference?
The three files do three different jobs, and none replaces another. robots.txt sets rules for bots, sitemap.xml lists your pages for search engines, and llms.txt is a short guide for AI tools.
A common mix-up is to use llms.txt to block or allow AI bots. It holds no rules, only text and links.
| What | llms.txt | robots.txt | sitemap.xml |
|---|---|---|---|
| Job | A short guide to the site, with links to its key pages | Rules that tell crawlers which URLs they may access | A file that tells search engines about the pages and files on your site |
| Written in | Markdown | Plain text rules such as User-agent and Disallow | XML |
| Where it lives | /llms.txt, or inside a folder such as /docs/llms.txt | /robots.txt | Usually /sitemap.xml |
| Can it block a bot? | No | Yes, for bots that follow it | No |
| Does Google Search use it? | No. Google says Search ignores it | Yes | Yes |
How do you add llms.txt on Shopify, WordPress and other platforms?
On most sites you upload a file named llms.txt to the root folder, the top level of the site, so it opens at yourdomain.com/llms.txt. Hosted platforms differ, and some create the file for you. This is what each platform's own documentation says.
Shopify
A Shopify store already answers at /llms.txt, so there is no file to create or upload. Shopify serves a file called agents.md for every store, and /llms.txt returns the same content by default. It is written for AI agents and includes the store name, store URL, sitemap link and store policies (Shopify Help Center, checked 7 Oct 2026).
Shopify says the managed file is all most stores need, and that in most cases you should not add a separate llms.txt.liquid template. To change the content, it points first to a theme template named agents.md.liquid, which changes /agents.md, /llms.txt and /llms-full.txt together. It keeps llms.txt.liquid for an advanced need, when /llms.txt must differ from /agents.md (Shopify developer docs and Shopify Help Center, checked 7 Oct 2026).
If you do need that template, these are Shopify's steps. In the Shopify admin, go to Online Store > Themes. Find your theme, click the "..." button and choose Edit code. In the Templates folder, create a new file named llms.txt.liquid. The template accepts plain text, Markdown and Liquid, Shopify's template language. From then on you edit it by hand and keep its content current yourself (Shopify developer docs, checked 7 Oct 2026).
The managed file holds instructions for agents, not a list of your pages, so the checker below shows warnings for it. That is expected, and it is no reason to replace the file. Our Shopify SEO guide covers the rest of a store's search setup.
WordPress
WordPress.com's support guide gives two routes. One is a plugin such as Yoast SEO or Rank Math. The other is manual. You write the file and upload it over SFTP, a secure way to move files, into the site's root folder (htdocs). WordPress.com offers this on its Business and Commerce plans (WordPress.com support, checked 8 Oct 2026).
Yoast says its plugin creates the file once you switch the feature on, at no extra cost (Yoast, checked 8 Oct 2026).
Wix and Squarespace
- Wix. Wix creates and maintains the file automatically. You can view it, edit it or switch it off under SEO & GEO in the dashboard. Wix lists three things to do first: upgrade the site, connect a custom domain and let search engines index it (Wix Help Center, checked 7 Oct 2026).
- Squarespace. The file is off by default. On version 7.1 sites you turn it on by adding text in the LLMS.txt tab of SEO Settings (Squarespace Help Center, checked 7 Oct 2026).
Any other site
If you control the server, put the file in the folder that holds your home page. Then open yourdomain.com/llms.txt in a browser. You should see plain text, not your page design or a 404 page.
What mistakes should you avoid with llms.txt?
Apart from using it to block bots, which it cannot do, these are the mistakes to check.
- The wrong name or place. A tool that asks for /llms.txt will not find a file named llm.txt, or one uploaded to a media or files folder.
- Relative links. Write full addresses that start with https://. A link such as /pricing has no meaning away from your site.
- Listing every page. The proposal describes the file as a curated overview, and your sitemap already lists every page (llmstxt.org, checked 7 Oct 2026). Pick the pages that explain what you sell, your prices and your policies.
- Skipping the test. The proposal suggests a simple test: give an AI assistant only your llms.txt and ask it questions about your site (llmstxt.org, checked 7 Oct 2026).
How do you check your llms.txt file?
Check it in two steps. First open yourdomain.com/llms.txt in a browser. You should see plain text that starts with a "# " title line, not your page design or a 404 page. Then select all the text, copy it and paste it into the checker on this page.
The checker runs nine checks on the format of the text you paste, and marks each one pass, warning or fail. It runs in your browser, and nothing you paste is uploaded. It does not test whether your site serves the file, or whether any AI engine reads or uses it.
- Title. There is one H1 title, a line that starts with "# ", and it is the first line.
- Summary. A blockquote summary, a line that starts with "> ", sits directly under the title.
- Sections. Link sections use H2 headings, lines that start with "## ", and nothing deeper.
- List items. Each list item under a section heading is a Markdown link.
- Addresses. Each link uses a full web address that starts with http:// or https://.
- Empty sections. No section is left without a link.
- Markdown. The text is Markdown, not an HTML page or an XML file.
- Not robots.txt. The text holds no robots.txt rules, such as User-agent or Disallow lines.
- Size. The text is no longer than 100,000 characters. That line is the checker's own, because the proposal sets no limit.
Where Depra AI fits
Depra AI is AI visibility tracking software. Its AI crawler checker reads your robots.txt for 41 AI crawlers from 26 vendors. It gives each one a verdict: allowed, partly allowed or blocked. The check sits behind a login and comes with every plan. See the crawlability docs for that report.
robots.txt matters more than llms.txt here. A missing llms.txt costs you little. A blocked search bot can keep you out of that engine's answers.
Depra AI does not write your llms.txt, upload it or change your website. The generator and checker on this page are free and need no account. For the wider work of getting named in AI answers, read our guide to AI SEO.