> For clean Markdown content of this page, append .md to this URL. For the complete documentation index, see https://buildwithfern.com/learn/llms.txt.

# Markdown access

> Access your documentation as Markdown — through per-page URLs and `llms.txt` — for AI agents and tooling.

Fern serves clean Markdown for any documentation page so agents can consume your content efficiently. Agents fetch the source by appending `.md` or `.mdx` to a page URL, or by sending an `Accept: text/markdown` header via [content negotiation](https://developer.mozilla.org/en-US/docs/Web/HTTP/Content_negotiation). Combined with [`llms.txt`](/learn/docs/ai-features/llms-txt), this reduces token consumption by 90%+ compared to HTML.

`.md` and `.mdx` URLs resolve for every routable page, including API Reference endpoint pages that don't appear in the navigation, `sitemap.xml`, or `llms.txt`. Any URL that renders as HTML returns its Markdown twin.

![Example showing a page's underlying Markdown](/learn/_fern-img/ddc59739cea5e45e26d177e58ed6ef3476660d76460d054a017859b43f71a978.webp)

The same Markdown is used everywhere — single pages and `llms.txt` — and respects the same `<llms-only>` and `<llms-ignore>` content controls.

A [default per-page directive](/learn/docs/ai-features/agent-directives) is automatically prepended to every page's Markdown output when served to AI agents, pointing them to your `.md` URLs and `llms.txt`. You can [override or disable this directive](/learn/docs/configuration/site-level-settings#agents-configuration) in `docs.yml`. The directive is only visible to agents — human-facing documentation is unaffected.

## API Reference pages

Endpoint and webhook pages render as a flattened property list. Fern emits a `## Request` section covering path parameters, query parameters, headers, and body properties, a `## Response` section with the response shape for each status code, and an `## Authentication` section for the endpoint's security schemes.

Each property renders as a bullet with its name, type, required flag, and description, with nested objects indented one level and types written in the same shorthand as the docs UI (`string`, `list of objects`, `map from string to integer`). Internal component schema names and `$ref` chains are omitted.

**`plants/update-plant.md`**

```md title="plants/update-plant.md"
## Authentication

- `Authorization` header (bearer token, required)

## Request

### Path parameters

- `plantId` (string, required) — The plant to update.

### Body (application/json)

- `name` (string, required) — Common name of the plant.
- `sunlight` (object, optional)
  - `hoursPerDay` (integer, required)

## Response

### 200

- `id` (string, required) — Unique identifier for the plant.
```

## Interactive components in Markdown output

Interactive components such as [`<EndpointRequestSnippet>`](/learn/docs/writing-content/components/request-snippet), [`<EndpointResponseSnippet>`](/learn/docs/writing-content/components/response-snippet), and [`<EndpointSchemaSnippet>`](/learn/docs/writing-content/components/endpoint-schema-snippet) render as fenced code blocks and structured content in the Markdown output, so agents receive the full request and response examples without parsing HTML.

## Redirects

The [`redirects` configured in `docs.yml`](/learn/docs/seo/redirects) apply to Markdown URLs. A redirected page's `.md` or `.mdx` twin returns a `308` to the destination's twin, preserving the requested extension: `/old/page.md` redirects to `/new/page.md`, and `/old/page.mdx` to `/new/page.mdx`.

## Missing pages

When an agent requests a `.md` URL that matches no page and no redirect, the response includes up to three suggested pages under a `## Similar pages` heading, each linking to the closest matching `.md` source. This mirrors the "Were you looking for one of these?" suggestions human readers see on the 404 page.

## Accessing protected docs

On sites with [authentication](/learn/docs/authentication/overview) enabled, agents must include a JWT on every Markdown request — whether for an individual page or `llms.txt`. Exchange your [Fern API key](/learn/cli-api-reference/cli-reference/general-commands#fern-token) for a [JWT](/learn/docs/fern-api-reference/get-jwt):

**`Get a JWT`**

```bash Get a JWT
curl https://docs.example.com/api/fern-docs/get-jwt \
  -H "FERN_API_KEY: $FERN_TOKEN"
# → { "fern_token": "eyJ...", "roles": [] }
```

Send the returned JWT as the `FERN_TOKEN` header on subsequent requests:

**`Fetch protected content`**

```bash Fetch protected content
curl https://docs.example.com/platform/overview \
  -H 'Accept: text/markdown' \
  -H "FERN_TOKEN: $JWT"
```

JWTs are valid for 30 days — cache and refresh as needed.

## Markdown for troubleshooting

Viewing the Markdown directly is also useful for troubleshooting layout problems. A **View as Markdown** button is enabled by default on every page and can be configured through the [page actions configuration](/learn/docs/configuration/site-level-settings#page-actions-configuration).

> Access your documentation as Markdown — through per-page URLs and `llms.txt` — for AI agents and tooling.