What Is llms.txt and How Do You Create One?
llms.txt is a Markdown file placed in a website's root directory that briefly describes the site to AI agents and links to its most important pages. It is not an official standard but a community proposal made in 2024, and Google Search does not use it. It is not required; it can be a low-cost addition, but it does not replace the fundamentals of visibility.
With the rise of AI search, llms.txt has started to come up often, and some sources present it as "robots.txt for AI" or as a key to visibility. The real picture is simpler: the file is easy to create, but what it achieves is limited and debated.
This article explains what llms.txt is, how to create one, an example file, and what you should and should not expect from it. For the wider context, see our article What Is GEO?
Quick answers about llms.txt
Question | Short answer |
|---|---|
What is it? | A Markdown file in the site's root directory that summarises the site and links to its important pages |
Who proposed it? | Jeremy Howard (Answer.AI), 3 September 2024; the proposal is published at llmstxt.org |
Is it an official standard? | No, it is a community proposal; version 2 was published in August 2026 |
Is it required? | No |
Does Google use it? | No. According to Google, Search, AI Overviews and AI Mode do not use this file |
Where does it go? | In the domain's root directory: yourdomain.com/llms.txt |
Who might benefit? | Mainly content sites read by AI agents, such as developer documentation |
What is llms.txt?
llms.txt is a plain Markdown file placed in a website's root directory. Its purpose is to briefly introduce the site to agents built on large language models and to link to its most important pages. The idea behind the proposal is this: HTML pages are full of navigation menus, ads and scripts, so a clean, curated "table of contents" can be useful to an agent.
The proposal was published by Jeremy Howard of Answer.AI on 3 September 2024 and lives at llmstxt.org. It has not been adopted by a standards body; it lives on as a community proposal, and its second version was published in August 2026.
Is llms.txt required, and does Google use it?
No. llms.txt is not required, and Google Search does not use it. Google's guide to generative AI features says that "special" files such as llms.txt are not needed to appear in Google Search, including AI Overviews and AI Mode. The same guide says it is fine to create and maintain such a file for other systems, but that doing so will neither harm nor help your visibility in Search.
Source: Google Search Central, Optimizing your website for generative AI features.
Do AI crawlers read llms.txt?
As far as we know, none of the major AI search products has documented that it reads the file or uses it for ranking or citation. Independent research points the same way: an SE Ranking study of around 300,000 domains found no relationship between having the file and AI citations, and some server log analyses show that AI bots request the file very rarely.
The exception is coding agents and developer documentation: some software companies publish their documentation indexes in this format. So the file does have a real use case, but for a corporate website it is not a "key to AI visibility".
The main tool for controlling crawler access is still robots.txt; OpenAI's crawler documentation also describes managing its bots through robots.txt. We summarised what each AI crawler does in our article Why Isn't Your Brand Showing Up in ChatGPT?
Where does llms.txt go?
In the domain's root directory: https://yourdomain.com/llms.txt. The file should be publicly accessible, require no login and respond to a normal request with a 200 status code. Version 2 of the proposal also allows the file to sit in a subdirectory (for example /docs/llms.txt), in which case it covers the pages under that path.
llms.txt example
The example below was written for a fictional company. The company name, addresses and descriptions are not real.
# Example Machinery Inc. > Example Machinery Inc. is a company that manufactures industrial packaging machines. This file is a short guide to the main sections of the site. Product information and support pages are listed below. ## Company - [About us](https://example.com/about): Company profile, history and certifications - [Contact](https://example.com/contact): Address and contact form ## Products - [Packaging machines](https://example.com/products/packaging): Product groups and general features - [Spare parts](https://example.com/products/spare-parts): Spare parts catalogue and ordering information ## Support - [Frequently asked questions](https://example.com/faq): Installation, maintenance and service questions ## Optional - [Press room](https://example.com/press): News and announcementsThe structure works like this: the site name comes first as an H1, followed by a short summary (as a blockquote), then lists in the form "link: short description" under H2 headings. The "Optional" section is for pages of secondary importance that can be skipped if needed.
How do you create llms.txt?
Decide. The file does nothing for Google; it only makes sense if you have a documentation-focused site or expect agent traffic. Because the cost is low, trying it is also reasonable.
Choose 10 to 20 core pages. Pages an agent needs to understand you, such as about us, service or product pages, FAQs, contact and certifications.
Write a one-sentence summary. Describe who you are and what you do in neutral, verifiable language; do not write a marketing slogan.
List the links with descriptions. Add a short note to each link saying what the page contains.
Upload it to the root directory. Publish the file at /llms.txt through your CMS or server.
Test it. Open it in a browser and check that it returns a 200 response and that every link works.
Keep it up to date. Update the file when pages change or are removed; it should stay consistent with the information on your site.
What is llms-full.txt?
llms-full.txt is a convention some sites publish alongside llms.txt, gathering the content of their pages into one large file instead of linking to them. The current text of the proposal does not define this file, so it is an unofficial, community-practice extension. A corporate website is unlikely to need it.
What should you prioritise before llms.txt?
Work with a proven effect on visibility comes first: verifying crawler access and CDN rules, putting a direct answer paragraph at the top of every important page, keeping company information consistent everywhere and matching structured data to the page. For the overall framework, see our article on SEO vs GEO vs AEO. llms.txt is an optional piece added on top of that foundation, not a replacement for it.
Clockwork's view from the field
Clockwork is an independent digital and creative agency based in Istanbul with 25 years of experience, certified to ISO 27001 and ISO 12207.
Frequently asked questions
What is llms.txt?
llms.txt is a Markdown file placed in a website's root directory that briefly describes the site to AI agents and links to its important pages. It is a community proposal made by Jeremy Howard in 2024, not an official standard.
Is llms.txt required?
No. Google says Search does not use this file and that creating one will neither harm nor help your visibility in Google Search. You are free to create one for other systems.
Where does llms.txt go?
In the domain's root directory, at yourdomain.com/llms.txt. The file should be publicly accessible and respond with a 200 status code.
Does llms.txt affect Google rankings?
No. Google says Search does not use llms.txt and that the file does not change visibility or rankings in Google Search, including AI Overviews and AI Mode.
Conclusion
llms.txt is easy to create, and expectations for it should be kept low. It is not required, Google does not use it, and its effect on AI visibility has not been proven. If you have a documentation-heavy site or a product that expects agent traffic, it is worth trying; for a corporate website, look first at the fundamentals such as crawler access, direct answers and consistent information.
Clockwork is an independent digital and creative agency based in Istanbul with 25 years of experience. If you would like to assess your site's AI visibility together, explore our services or get in touch.