---
title: "htmlToMarkdown · Affordances"
description: "The converter on its own, with no React or Next.js dependency."
url: "https://affordances.vercel.app/docs/plugins/md/converter"
image: "https://affordances.vercel.app/og/docs/plugins/md/converter"
lang: "en"
---

# htmlToMarkdown

The converter on its own, with no React or Next.js dependency.

`htmlToMarkdown` from `@affordances/md` is the converter the route handler uses. Give it any HTML document:

```ts
import { htmlToMarkdown } from '@affordances/md'

const html = await (await fetch('https://example.com/post')).text()
const { markdown, title, url, tokens } = htmlToMarkdown(html, { url: 'https://example.com/post' })
```

| Option        | Default  | Description                                                                                                       |
| ------------- | -------- | ----------------------------------------------------------------------------------------------------------------- |
| `url`         | none     | Base for absolute links and images, usually the page's URL                                                        |
| `extract`     | `'auto'` | `'auto'`: markers, then `<main>`, then the body minus chrome; `'none'`: the whole body; or a function of the tree |
| `frontmatter` | `true`   | Prepend YAML frontmatter from the page's metadata                                                                 |
| `jsonLd`      | `true`   | Append the page's JSON-LD as fenced JSON                                                                          |
| `handlers`    | none     | Per-element overrides, as `hast-util-to-mdast` handlers                                                           |

It returns the Markdown, the page's `title`, its canonical `url` (or the one you passed), and `tokens`, an estimate of the Markdown's length in tokens (characters divided by four).