Merge, split, render and OCR PDFs directly in your n8n workflows — developed by prokodo.

✨ Features

• 🧩 Merge multiple PDFs into one
• ✂️ Split a PDF into pages or custom ranges
• 🖼️ PDF → Images (PNG/JPEG) at a chosen DPI
• 🔎 OCR (Tesseract.js) on PDFs or images
• 🤖 Auto-detects PDF vs image input
• 📑 Per-page or combined text output
• 📎 Optional TXT attachment

✅ Requirements

• Node.js 18+ / 20 LTS
• n8n ≥ 1.103 (tested on 1.105+)

Using an older n8n (e.g. 1.88)? It may still work if you align n8n-core / n8n-workflow versions. For best results, upgrade n8n.

📦 Install

Option A — Install into your n8n “custom extensions” folder (recommended)

Local n8n (not Docker):

# choose your custom folder (default ~/.n8n)
export N8N_CUSTOM_EXTENSIONS=~/.n8n

# install the node into that folder
npm install --prefix "$N8N_CUSTOM_EXTENSIONS" @prokodo/n8n-nodes-pdf-toolkit@latest

# start n8n
n8n start

Docker (example Dockerfile):

FROM n8nio/n8n:latest

ENV N8N_CUSTOM_EXTENSIONS=/home/node/.n8n
ENV NODE_PATH=/home/node/.n8n/node_modules

USER node
RUN npm install --prefix /home/node/.n8n @prokodo/n8n-nodes-pdf-toolkit@latest

After starting n8n, search in the node picker for “prokodo (PDF Toolkit)” Internal name: prokodoPdfToolkit

🛠 Dev install (build + link locally)

# in this repo
npm ci
npm run build

# make your package linkable
npm link

# link into your n8n custom extensions folder
npm link @prokodo/n8n-nodes-pdf-toolkit --prefix ~/.n8n

# start n8n with your custom folder
export N8N_CUSTOM_EXTENSIONS=~/.n8n
n8n start

Publish-ready tip: This package publishes compiled JS from dist/ to npm. You don’t need to commit dist/ to Git. To support installs straight from GitHub, add:

"scripts": {
  "prepare": "npm run build"
}

…and commit src/ (not dist/).

🔎 Node usage

Common

• Binary Property (Input): name of the incoming binary property holding your PDF/image (default data).

Merge PDFs

Operation: Merge PDFs

• Output Property (Merged PDF): binary property for output (default data)
• Output File Name: e.g. merged.pdf
• Result: a single item with the merged PDF.

Split PDF

Operation: Split PDF

• Split Mode: Every Page or By Ranges
• Ranges (if By Ranges): e.g. 1,3-5,10- (1-based, inclusive; 10- means page 10 to end)
• Output Property: binary property for each part (default data)
• Result: one item per output part.

PDF → Images

Operation: PDF to Images

• Image Format: PNG or JPEG
• DPI: e.g. 150 (higher → bigger & slower)
• JPEG Quality: only when format is JPEG
• Page Ranges: e.g. 1-2,5
• Result: one item per rendered page, each with binary.image.

OCR

• Operation: OCR
• Languages (Tesseract): e.g. eng, deu, eng+deu
• Return: Single (combined text) or Per Page
• Also Attach TXT as Binary: optional .txt file output

Advanced Settings (optional):

• OCR Page Ranges (for PDFs only)
• OCR Image Format, OCR DPI, OCR JPEG Quality (rasterization settings)
• Tesseract Lang Path (custom URL/path hosting *.traineddata)
• OCR auto-detects whether the incoming file is a PDF or image via MIME/bytes.

OCR auto-detects whether the incoming file is a PDF or image via MIME/bytes.

🧯 Troubleshooting

Node doesn’t show up

Ensure N8N_CUSTOM_EXTENSIONS points to the folder where you installed the package. Restart n8n and search for “prokodo (PDF Toolkit)”. Verify your n8n version (≥ 1.103 recommended).

“Could not get parameter” during execution

Open the node in the editor and ensure all visible fields for the chosen operation are set. If you switched operations, re-open and re-save the node. canvas native dependency errors

On Debian/Ubuntu containers:

apt-get update && apt-get install -y \
  libcairo2 libpango-1.0-0 libjpeg62-turbo libgif7 librsvg2-2 \
  && rm -rf /var/lib/apt/lists/*

On Alpine:

apk add --no-cache cairo pango jpeg giflib librsvg

Prefer Debian-based n8n images for smoother canvas support.

OCR language data not found / offline

• Set Tesseract Lang Path to a URL/path hosting *.traineddata, e.g. https://tessdata.projectnaptha.com/4.0.0
• Use ocrLang like eng, deu, or eng+deu.

Slow/High memory on large PDFs

• Lower DPI (e.g. 120).
• Use page ranges (process in chunks).
• Avoid very high DPI for OCR unless necessary.

🙌 Contributing

PRs welcome!

npm ci
npm run build

Open a PR with what changed and how to test it.

📄 License

This library is published under MIT.

© 2025 prokodo. Visit us at prokodo.com.