> This page location: Community > Using docs as Markdown (LLMs)
> Full Neon documentation index: https://neon.com/docs/llms.txt

# Getting Neon docs as Markdown

How to get our documentation as plain Markdown for LLMs, tools, and scripts

Neon provides its documentation in plain Markdown so that LLMs, API clients, and scripts can consume it easily. We follow the [llms.txt](https://llmstxt.org/) approach: same URLs as the website, with Markdown available when requested.

## How to get Markdown

You can get Markdown by requesting it explicitly: use the same docs URL with an `Accept` header, or append `.md` to the path.

1. **Same URL, request Markdown:** Use the same path you would use in a browser (for example, `/docs/connect/choose-connection`). Send an `Accept: text/markdown` or `Accept: text/plain` header, or make the request with a user agent string associated with ChatGPT, Claude, Cursor, etc. We respond with the Markdown version of that page. For example:

   ```bash
   curl -H "Accept: text/markdown" https://neon.com/docs/connect/choose-connection
   ```

2. **Append `.md` to the path:** Request the URL with `.md` at the end. For example:
   - HTML: [https://neon.com/docs/connect/choose-connection](https://neon.com/docs/connect/choose-connection)
   - Markdown: [https://neon.com/docs/connect/choose-connection.md](https://neon.com/docs/connect/choose-connection.md)

Whether you use the `Accept` header or the `.md` extension, you get the same Markdown output.

### Why not serve the source MDX?

Our docs are authored in MDX (Markdown plus components) and stored in the [neon-website](https://github.com/neondatabase/website/tree/main/content/docs) repository. If you read those source files directly, custom components and shared snippets can obscure the content or leave it incomplete (for example, a component might inject text that does not appear inline in the file).

The Markdown we serve is built from that MDX and gives you a single, flat page with all content in place. This is why we serve plain Markdown over the raw MDX.

## Finding content: the index

A full table of contents is available at:

- [https://neon.com/docs/llms.txt](https://neon.com/docs/llms.txt) (canonical)
- [https://neon.com/llms.txt](https://neon.com/llms.txt) (also serves the same file)

The index lists our doc pages with titles and links. Use it to discover URLs or to feed a list of pages into your tool.

## What you get on each page

Each Markdown page is plain Markdown (no interactive components or JSX). We add a bit of structure so that consumers know where they are and what to read next:

- **At the top:** A short block with the page location in the doc tree (e.g. `Connect > Choose a connection`) and a link to the full index (llms.txt). That gives context and a way to jump to the full catalog.

- **Body:** The main content, converted from our source format into standard Markdown (headings, links, code blocks, lists, etc.).

- **At the bottom:** A "Related docs" section with links to sibling pages in the same section, which is visible in the HTML version in the navigation bar. That helps with discovery and follow-up reading.

Each URL gives you the page content plus lightweight navigation and discovery.

## Headers and discovery metadata

We set a number of headers on our doc responses. These are the ones relevant to Markdown discovery and retrieval:

- `<link rel="alternate" type="text/markdown" href="...">`: Each HTML doc page includes this tag in the head, so agents can discover the Markdown URL without knowing about `.md` or `Accept` header conventions.
- `X-Robots-Tag: noindex`: Markdown responses include this header so search engines don't index them alongside the HTML versions.
- `X-LLMs-Txt` and `Link: rel="llms-txt"`: HTML doc responses include these headers pointing to `/docs/llms.txt` (an index of all docs as Markdown URLs), so agents can find the index from any page without parsing HTML. We adopted this convention from [Mintlify](https://mintlify.com).

## Summary for agents

| Goal                        | What to do                                                                                                     |
| --------------------------- | -------------------------------------------------------------------------------------------------------------- |
| Get one page as Markdown    | Request the HTML docs URL with `Accept: text/markdown` or append `.md` to the URL                              |
| Get the full list of docs   | [https://neon.com/docs/llms.txt](https://neon.com/docs/llms.txt)                                               |
| Understand page context     | Read the short block at the top of each Markdown page (location + index link)                                  |
| Find related pages          | Use the "Related docs" section at the bottom of each page                                                      |
| Discover Markdown from HTML | Look for `<link rel="alternate" type="text/markdown" href="...">` in the page head, or the `X-LLMs-Txt` header |

---

## Related docs (Community)

- [Community hub](https://neon.com/docs/community/community-intro)
- [Docs contribution guide](https://neon.com/docs/community/contribution-guide)
- [Using Mermaid diagrams](https://neon.com/docs/community/mermaid-diagrams)
- [Component guide](https://neon.com/docs/community/component-guide)
- [Component specialized guide](https://neon.com/docs/community/component-specialized)
- [Component icon guide](https://neon.com/docs/community/component-icon-guide)
- [Component architecture](https://neon.com/docs/community/component-architecture)
- [AI tools for documentation](https://neon.com/docs/community/ai-tools)