n8n-nodes-docx

Convert documents between Word (DOCX), Excel (XLSX), Markdown (MD), PDF, HTML, CSV and plain text — as an n8n community node, and as a reusable Node.js library.

100% pure JavaScript — no LibreOffice, no system binaries, no setup. Works on any n8n instance (including managed/Docker hosts) the moment you install it.

Conversion matrix

| From \ To | PDF | DOCX | XLSX | CSV | MD | HTML | TXT | |-----------|:---:|:----:|:----:|:---:|:---:|:----:|:---:| | DOCX | ✅ | — | — | — | ✅ | ✅ | ✅ | | XLSX | ✅ | — | ✅ | ✅ | ✅ | ✅ | ✅ | | CSV | ✅ | — | ✅ | — | ✅ | ✅ | ✅ | | MD | ✅ | ✅ | — | — | — | ✅ | ✅ | | HTML | ✅ | ✅ | — | — | ✅ | — | ✅ | | PDF | — | ⚠️ | ⚠️ | ⚠️ | ⚠️ | ⚠️ | ✅ |

✅ supported · ⚠️ best-effort (text-layer extraction; layout/images not preserved) · — not meaningful

Fidelity notes

Pure-JS conversion favours content and structure (text, headings, lists, tables, bold/italic) over pixel-perfect layout. Complex page layouts, embedded fonts and images are simplified. For most Word/Excel/Markdown/PDF automation this is exactly what you want; if you need print-perfect fidelity, a LibreOffice-backed service is the alternative.

Requirements

• Node.js ≥ 18.10

That's it. All conversion libraries ship inside the package.

Install (n8n)

In n8n: Settings → Community Nodes → Install, enter n8n-nodes-docx.

Or manually:

cd ~/.n8n/nodes        # or your N8N_CUSTOM_EXTENSIONS dir
npm install n8n-nodes-docx

Add a DocX Convert node, pick Source Format (or Auto-Detect) and Target Format, and point it at the binary property holding your file (default data).

Use as a library

import { convert } from 'n8n-nodes-docx';
import { readFile, writeFile } from 'node:fs/promises';

const docx = await readFile('report.docx');
const { data, ext, mime } = await convert(docx, { from: 'docx', to: 'pdf' });
await writeFile(`report.${ext}`, data);

// Markdown → Word
const pdf = await convert('# Hello\n\nWorld', { from: 'md', to: 'docx' });

// Auto-detect from filename
await convert(buf, { to: 'md', filename: 'sheet.xlsx' });

API

convert(input: Buffer | string, opts: {
  from?: 'docx'|'xlsx'|'md'|'pdf'|'html'|'txt'|'csv'; // auto-detected if omitted
  to:    'docx'|'xlsx'|'md'|'pdf'|'html'|'txt'|'csv';
  filename?: string;     // helps auto-detection
}): Promise<{ data: Buffer; mime: string; ext: Format }>

Development

npm install
npm run build      # tsc → dist/
npm test           # node --test (real conversions, no system deps)
npm run dev        # tsc --watch

Project layout:

src/                 reusable conversion library
  convert.ts         routing / orchestration
  readers.ts         docx/xlsx/pdf/html → normalized form
  writers.ts         → pdf (pdfkit), docx, xlsx, csv, md, html
  html-model.ts      HTML → structured blocks (shared by pdf/docx writers)
  markdown.ts        md ↔ html (marked / turndown)
  formats.ts         format detection + MIME
nodes/Docx/          n8n community node + icon
test/                conversion tests

Powered by: pdfkit, docx, xlsx (SheetJS), mammoth, pdf-parse, marked, turndown, html-to-text — all pure JS.

License

MIT