n8n-nodes-german-ocr

Community node for n8n to interact with the German-OCR API.

Provides OCR text extraction, invoice data extraction, document anonymization (Privacy Shield) and semantic document search directly inside your n8n workflows.

Installation

In n8n (Community Nodes)

1. Open Settings > Community Nodes in your n8n instance.
2. Click Install a community node.
3. Enter n8n-nodes-german-ocr and click Install.

Manual / npm

cd ~/.n8n/custom
npm install n8n-nodes-german-ocr
# Restart n8n

Docker

Add to your Dockerfile:

RUN cd /usr/local/lib/node_modules/n8n && npm install n8n-nodes-german-ocr

Credentials

| Field | Description | | ----------- | ----------- | | API Key | Your German-OCR API key from app.german-ocr.de/settings | | API Secret | Your German-OCR API secret | | Base URL | https://api.german-ocr.de (default) |

Operations

Document

| Operation | Description | | --- | --- | | Analyze | Upload a document for OCR. Choose between Turbo, Pro, Ultra and Max models. Polls until the job is complete and returns the extracted text, pages, fields and confidence score. | | Extract Invoice | Shortcut that uses the Ultra model optimised for structured invoice data extraction (sender, recipient, line items, totals, etc.). | | Anonymize | Uses the Privacy Shield model to detect and redact personally identifiable information (PII). |

Search

| Operation | Description | | --- | --- | | Query | Perform a semantic (meaning-based) search across all indexed documents. | | Index Document | Upload a document to the semantic search index. | | List Documents | Retrieve a list of all currently indexed documents. |

Usage Examples

Basic OCR Workflow

1. Add a Read Binary File node (or any node that outputs binary data).
2. Connect it to the German-OCR node.
3. Set Resource = Document, Operation = Analyze, choose a model.
4. The output contains result.text, result.pages, result.fields and result.confidence.

Invoice Processing

1. Trigger via email attachment or file upload.
2. Use German-OCR > Document > Extract Invoice.
3. Map the structured JSON output to your accounting system.

Semantic Search

1. First index documents with Search > Index Document.
2. Later query with Search > Query using natural language.

Models

| Model | Speed | Accuracy | Best for | | --- | --- | --- | --- | | german-ocr-turbo | Fastest | Good | Simple, clean documents | | german-ocr-pro | Fast | High | General-purpose OCR | | german-ocr-ultra | Medium | Very High | Invoices, complex layouts | | german-ocr-max | Slow | Highest | Difficult scans, old documents | | privacy-shield | Medium | – | PII detection & redaction |

Development

# Clone the repo
git clone https://github.com/german-ocr/n8n-nodes-german-ocr.git
cd n8n-nodes-german-ocr

# Install dependencies
npm install

# Build
npm run build

# Link into local n8n for testing
npm link
cd ~/.n8n/custom && npm link n8n-nodes-german-ocr

License

MIT