llms.txt, explained

What llms.txt is

llms.txt is a Markdown file at the root of a website, at /llms.txt, that tells AI agents what the site is and where its most useful pages are. It was proposed by Jeremy Howard in September 2024; the current version of the proposal was updated in August 2026.

The idea is simple. Web pages are built for people, wrapped in navigation, scripts and layout. An agent with a limited context window does better with a short, plain summary and a list of links to the pages that answer real questions.

The format

The proposal defines the sections, in this order:

1. An optional byte-order mark.
2. An H1 with the name of the project or site. This is the only required part.
3. A blockquote with a short summary.
4. Optional Markdown without headings: paragraphs or lists that help an agent read the links.
5. H2 sections, each a list of links written as - [Title](URL): notes.

A section named Optional holds links an agent can skip when it is short of context.

# LogZuno

> Event tracking and product analytics with web analytics that separate real people from bots.

## Docs

- [Quickstart](https://logzuno.app/docs/quickstart): Send your first event
- [Ingestion API](https://logzuno.app/docs/api/ingestion): Endpoints, limits and errors

## Optional

- [Blog](https://logzuno.app/blog)

What changed in version 2

The 2026 revision makes three additions worth knowing:

• Files below the root. An llms.txt can sit at any path, such as /docs/llms.txt, and covers the pages under it. Where several apply, agents should use the most specific.
• Markdown copies of pages. Pages can offer a clean Markdown version at the same URL with .md added.
• Link relations. rel="alternate" type="text/markdown" points to a page's Markdown version and rel="describedby" points to the llms.txt that covers it, as HTML links or a Link header.

The proposal notes that thousands of sites publish the file, that documentation platforms generate it, and that OpenAI, Anthropic and Google's Gemini publish llms.txt files for their own developer docs.

Who reads it

Mostly agents that go looking for it: coding assistants fetching a library's documentation, and chat assistants reading a site to answer a question. It isn't a search ranking signal, and robots.txt still decides what crawlers may fetch.

You can see whether it is read: requests for /llms.txt show up in your server logs, and LogZuno's AI visibility reports every fetch of it by agent.

Writing a good one

• Lead with one plain sentence about what you are and who it's for. Agents read the summary first.
• Link to pages that answer questions: pricing, product, docs, policies. Not every page in your sitemap.
• Write notes for each link that say what the page answers.
• Use absolute URLs, so the file still works when it's copied.
• Keep it current. A link to a page that no longer exists is worse than no link.

Build one with the free llms.txt generator and check an existing file with the llms.txt checker.