npm install @tpmjs/tools-html-to-markdown
# or
pnpm add @tpmjs/tools-html-to-markdown
# or
yarn add @tpmjs/tools-html-to-markdown

import { htmlToMarkdownTool } from '@tpmjs/tools-html-to-markdown';
import { generateText } from 'ai';

const result = await generateText({
  model: yourModel,
  tools: {
    htmlToMarkdown: htmlToMarkdownTool,
  },
  prompt: 'Convert this HTML to markdown',
});

import { htmlToMarkdownTool } from '@tpmjs/tools-html-to-markdown';

const result = await htmlToMarkdownTool.execute({
  html: '<h1>Hello World</h1><p>This is a <strong>test</strong>.</p>',
});

console.log(result.markdown);
// # Hello World
//
// This is a **test**.

console.log(result);
// {
//   markdown: '# Hello World\n\nThis is a **test**.',
//   wordCount: 6
// }

{
  headingStyle?: 'setext' | 'atx';      // Default: 'atx'
  bulletListMarker?: '-' | '*' | '+';   // Default: '-'
}

{
  markdown: string;    // The converted markdown
  wordCount: number;   // Number of words in the markdown
}

const result = await htmlToMarkdownTool.execute({
  html: '<h1>Title</h1><p>Paragraph text.</p>',
});

console.log(result.markdown);
// # Title
//
// Paragraph text.

console.log(result.wordCount);
// 3

const result = await htmlToMarkdownTool.execute({
  html: '<a href="https://example.com">Click here</a> for more info.',
});

console.log(result.markdown);
// [Click here](https://example.com) for more info.

const result = await htmlToMarkdownTool.execute({
  html: `
    <ul>
      <li>First item</li>
      <li>Second item</li>
      <li>Third item</li>
    </ul>
  `,
});

console.log(result.markdown);
// - First item
// - Second item
// - Third item

const result = await htmlToMarkdownTool.execute({
  html: '<ul><li>Item 1</li><li>Item 2</li></ul>',
  options: {
    bulletListMarker: '*',
  },
});

console.log(result.markdown);
// * Item 1
// * Item 2

const result = await htmlToMarkdownTool.execute({
  html: '<h1>Main Title</h1><h2>Subtitle</h2>',
  options: {
    headingStyle: 'setext',
  },
});

console.log(result.markdown);
// Main Title
// ==========
//
// Subtitle
// --------

const result = await htmlToMarkdownTool.execute({
  html: `
    <table>
      <thead>
        <tr><th>Name</th><th>Age</th></tr>
      </thead>
      <tbody>
        <tr><td>Alice</td><td>25</td></tr>
        <tr><td>Bob</td><td>30</td></tr>
      </tbody>
    </table>
  `,
});

console.log(result.markdown);
// | Name | Age |
// | --- | --- |
// | Alice | 25 |
// | Bob | 30 |

const result = await htmlToMarkdownTool.execute({
  html: '<pre><code>function hello() {\n  console.log("Hello");\n}</code></pre>',
});

console.log(result.markdown);
// ```
// function hello() {
//   console.log("Hello");
// }
// ```

const result = await htmlToMarkdownTool.execute({
  html: '<p>Use the <code>console.log()</code> function.</p>',
});

console.log(result.markdown);
// Use the `console.log()` function.

const result = await htmlToMarkdownTool.execute({
  html: '<img src="photo.jpg" alt="A beautiful sunset">',
});

console.log(result.markdown);
// ![A beautiful sunset](photo.jpg)

const result = await htmlToMarkdownTool.execute({
  html: '<p>This is <strong>bold</strong> and <em>italic</em> text.</p>',
});

console.log(result.markdown);
// This is **bold** and *italic* text.

const result = await htmlToMarkdownTool.execute({
  html: '<blockquote><p>This is a quote.</p></blockquote>',
});

console.log(result.markdown);
// > This is a quote.

const result = await htmlToMarkdownTool.execute({
  html: `
    <article>
      <h1>Blog Post Title</h1>
      <p>Introduction paragraph with <a href="/link">a link</a>.</p>
      <h2>Section 1</h2>
      <ul>
        <li>Point one</li>
        <li>Point two</li>
      </ul>
      <h2>Section 2</h2>
      <p>More content with <strong>bold</strong> and <em>italic</em>.</p>
      <pre><code>const x = 42;</code></pre>
    </article>
  `,
});

console.log(result.markdown);
// # Blog Post Title
//
// Introduction paragraph with [a link](/link).
//
// ## Section 1
//
// - Point one
// - Point two
//
// ## Section 2
//
// More content with **bold** and *italic*.
//
// ```
// const x = 42;
// ```

console.log(result.wordCount);
// 20

const result = await htmlToMarkdownTool.execute({
  html: '<div><span>Just plain text</span></div>',
});

console.log(result.markdown);
// Just plain text

const result = await htmlToMarkdownTool.execute({
  html: '<p>The **quick** brown fox.</p>',
});

console.log(result.wordCount);
// 4 (counts "The", "quick", "brown", "fox")

try {
  const result = await htmlToMarkdownTool.execute({
    html: null,  // Invalid input
  });
} catch (error) {
  console.error(error.message);
  // "HTML input must be a string"
}

const scrapedHtml = await fetch('https://example.com').then(r => r.text());
const result = await htmlToMarkdownTool.execute({ html: scrapedHtml });
// Now you have clean markdown for processing

const editorHtml = richTextEditor.getHTML();
const result = await htmlToMarkdownTool.execute({ html: editorHtml });
// Store as markdown instead of HTML

const emailHtml = getEmailBodyHtml();
const result = await htmlToMarkdownTool.execute({ html: emailHtml });
// Clean markdown representation of email

const apiDocsHtml = generateApiDocs();
const result = await htmlToMarkdownTool.execute({ html: apiDocsHtml });
// Markdown docs for GitHub/GitLab