llms.txt
llms.txt: llms.txt is a Markdown file placed at the root of a website (/llms.txt) that gives language models a summary of the site and a list of links to its most important content. Jeremy Howard proposed it in 2024.
llms.txt is a plain-text file in Markdown format, published at yourdomain.com/llms.txt. Its job is to give language models a clean, condensed version of your site: who you are, what you offer and where the important information lives. Jeremy Howard, co-founder of Answer.AI and fast.ai, proposed it in September 2024, and many technical documentation sites and companies working on AI search visibility have adopted it.
Why does llms.txt exist?
A modern web page is built for people: menus, cookie banners, scripts, ads and layout. To a language model, all of that is noise. Models also work with a limited context window and cannot read a whole site at once.
llms.txt offers a shortcut: a single structured plain-text file that explains the site and links to the pages worth reading. It is like handing someone an index instead of leaving them lost in the library.
What does an llms.txt file look like?
The proposal defines a simple Markdown structure:
- A
#title with the name of the site or company. - A short summary in a blockquote (
>), with the essentials. - Optional paragraphs with more context.
##sections that group lists of links, each with a short description.- An optional section called “Optional” for secondary links a model can skip when space runs short.
A minimal example for a clinic:
# Example Dental Clinic
> Dental clinic in Austin specializing in implants and clear aligners. Free first visit.
## Services
- [Dental implants](https://example.com/implants): price, process and timeline
- [Clear aligners](https://example.com/aligners): types and financing
## Practical information
- [Hours and location](https://example.com/contact)
Many sites also publish llms-full.txt, with the full text of the key pages in one file.
What does a business use llms.txt for?
For a business, llms.txt is a small piece of a GEO strategy, the work of getting assistants like ChatGPT, Perplexity or Gemini to find and cite you. It helps you:
- Control the summary: you decide how your company is described and which services appear first.
- Point to the good pages: guides, pricing, FAQs and service pages, instead of legal pages or tag archives.
- Make life easier for agents and tools that read websites to answer users, such as coding assistants or AI browsers.
Example: a B2B software company
Picture a company that sells management software for auto repair shops. Its site has 300 pages across blog, documentation, pricing and use cases. When someone asks an AI assistant “which shop management software supports online booking?”, the company wants the assistant to find the right page and not an old blog post.
The company publishes an llms.txt with a product summary, links to pricing, features and supplier integration docs, and puts the blog in the optional section. It also generates an llms-full.txt with the complete documentation. It keeps both updated automatically whenever it publishes a new page.
Is it worth creating?
Yes, with realistic expectations. Creating and maintaining the file costs little, especially if your CMS generates it automatically. But it does not replace what weighs most: a crawlable site, AI crawlers not blocked in robots.txt and every page answering one specific question clearly.
You can check whether your site has llms.txt, along with other technical points, with our AI visibility checker. If you want to work on it in depth, it is part of our generative engine optimization service.