REST API
Simple HTTP calls with JSON responses. Integrate within minutes in any programming language.
99%+ Accuracy
EasyData OCR Engine under the hood. European technology, refined through 25+ years of experience.
GDPR-compliant
European data centers, EU data processing. Documents are deleted immediately after processing.
European Support
Personal technical assistance. No chatbots, real developers who think along with you.
What is an OCR API?
An OCR (Optical Character Recognition) API enables you to extract text from scanned documents, images, and PDF files through a programmable interface. Instead of manual data entry, you automate the entire process from document upload to structured output.
The EasyData OCR API combines our powerful proprietary OCR engine with advanced machine learning models for optimal recognition of business documents. The result: accurate text extraction that’s immediately usable in your application.
How does our OCR API work?
You send a document (image, PDF, or scan) to our API endpoint. Our engine analyzes the document, recognizes text and structure, and returns the extracted data in JSON format. Optionally, we can extract specific fields such as invoice amounts, VAT numbers, or dates.
# Python example - OCR API call import requests # Upload and process document response = requests.post( "https://api.easydata.nl/v1/ocr", headers={"Authorization": "Bearer YOUR_API_KEY"}, files={"document": open("invoice.pdf", "rb")} ) # Receive structured data data = response.json() print(data["extracted_text"]) print(data["fields"]["invoice_total"])
API Capabilities
Our OCR API is the foundation — but EasyData offers developers a complete platform for intelligent document processing. From text extraction to advanced data analysis.
📄 Text Extraction
- Full-text OCR (all text)
- Table recognition with structure
- Handwriting recognition (ICR)
- Barcode & QR code scanning
- Multi-language support (50+ languages)
- ALTO XML output with coordinates
🎯 Field Extraction
- Invoice amounts & VAT
- Dates & references
- IBAN & company registration numbers
- Address data
- Custom field training
- Confidence scores
⚙️ Processing
- Automatic rotation
- Deskewing & denoising
- PDF/A conversion
- Batch processing
- Async webhooks
- Searchable PDF output
🚀 More than OCR: the complete EasyData platform
The OCR API is your starting point. Build further with specialized modules for complex document workflows.
📐 ALTO XML Export
Pixel-precise coordinates per word and line. Ideal for archiving, highlighting, and document reconstruction. More about ALTO →
🔍 Financial Search
Search millions of financial documents for specific amounts, IBANs, or transactions. Full-text search on steroids. More about Financial Search →
🤖 AI Document Intelligence
Combine OCR with machine learning for automatic classification, entity extraction, and smart routing. More about AI →
☁️ Nextcloud Integration
Seamlessly connect the EasyData OCR API to your Nextcloud environment. Automate document processing directly from your own cloud.
🔗 Why Nextcloud + OCR API?
Nextcloud provides a secure, self-hosted cloud solution. By connecting our OCR API via WebDAV and OCS interfaces, you create a fully automated document workflow:
- ✓ Automatic processing — Upload to a folder, receive OCR results back
- ✓ Data stays in your control — Processing via European servers, no US cloud
- ✓ Webhook triggers — Start OCR automatically with new files
- ✓ Searchable PDFs — Processed documents immediately searchable in Nextcloud
- ✓ Metadata tagging — Add extracted fields as tags
Ideal for organizations that take GDPR compliance seriously and want to keep their document processing fully under their own control.
Example: Nextcloud → OCR → Return
# Python: Fetch document from Nextcloud, OCR, return import requests # 1. Download document via WebDAV nc_url = "https://cloud.yourdomain.com/remote.php/dav/files/user/" doc = requests.get( nc_url + "Inbox/invoice.pdf", auth=("user", "app-password") ) # 2. Send to EasyData OCR API ocr_result = requests.post( "https://api.easydata.nl/v1/ocr", headers={"Authorization": "Bearer API_KEY"}, files={"document": doc.content} ) # 3. Upload searchable PDF back to Nextcloud requests.put( nc_url + "Processed/invoice_ocr.pdf", data=ocr_result.json()["searchable_pdf"], auth=("user", "app-password") )
Nextcloud APIs you’ll use
- ✓ WebDAV — Upload, download, move files
- ✓ OCS Sharing API — Manage shares after processing
- ✓ Capabilities API — Detect available features
Supported Documents
Invoices
Purchase invoices, sales invoices, credit notes. Automatic extraction of amounts, VAT, and supplier data.
Identity Documents
Passports, ID cards, driver’s licenses. MRZ reading and photo extraction with verification.
Contracts
Employment agreements, rental contracts, terms and conditions. Clause identification and metadata extraction.
Receipts
Cash register receipts, parking tickets, fuel receipts. Even crumpled and faded receipts are recognized.
Shipping Documents
CMR documents, packing slips, customs forms. Logistics document processing optimized.
Medical Documents
Prescriptions, lab results, referral letters. extraction with privacy-first approach.
Bank Statements
Account overviews, payment confirmations. Transaction parsing and accounting integration.
Handwritten
Forms, notes, questionnaires. for handwriting recognition.
Technical Specifications
Input Formats
PDF, TIFF, JPEG, PNG, BMP, GIF. Multi-page documents are automatically processed. Max 50MB per request.
Output Formats
JSON (default), XML, CSV, Excel, searchable PDF. Choose the format that fits your workflow.
Response Time
Average <2 seconds per page. Async processing available for large batches with webhook notifications.
Authentication
Bearer token authentication. API key management via dashboard. Rate limiting configurable per plan.
Integrations
SDKs for Python, Node.js, PHP, C#. Zapier connector available. Nextcloud, SharePoint, and custom webhooks.
SLA & Uptime
99.5% uptime guarantee. European servers with geographically separated backup. 24/7 monitoring.
Ready to integrate?
Technical fit? Let’s talk.
🚀 What to expect
API key on request Get started after request
Comprehensive documentation Code examples from experince
Personal onboarding Technical help with integration
European support Real developers who think along with you
Frequently Asked Questions
Which OCR engine do you use?
We use our own EasyData OCR Engine, developed over 25+ years. Our technology is specifically optimized for business documents and company-specific fields with machine learning models that continuously improve. Do you have specific requirements that can be best fulfilled through external technology? We also offer technical solutions for that, for example with ABBYY OCR.
What about the privacy of my documents?
Documents are processed on servers in the Netherlands and deleted immediately after processing. We do not store copies unless you explicitly request it. All connections are TLS 1.3 encrypted. We are GDPR-compliant and can sign a data processing agreement.
Can I test the API before paying?
Yes, of course, no obligations. You have access to all functionality, only with lower rate limits. Perfect for testing whether our API fits your use case.
How accurate is the OCR for business documents?
For printed text, we achieve 99%+ accuracy on well-readable documents. For handwriting, this varies between 85-95% depending on readability. Business-specific elements such as IBAN, company registration numbers, and VAT numbers are checked with extra validation.
How do I integrate with Nextcloud?
Integration works via Nextcloud’s WebDAV interface for file access and our REST API for OCR processing. You can automatically fetch documents from an “Inbox” folder, have them processed, and place the searchable PDF back. We provide sample code in Python, PHP, and Node.js. For Enterprise customers, we also offer a ready-made Nextcloud App.
Do you support batch processing?
Yes, you can upload multiple documents at once. For large batches (100+ documents), we offer async processing with webhook notifications. You receive a job ID and are notified when processing is complete.
Can I have custom fields trained?
Certainly, custom field extraction is possible based on your requirements. Send us sample documents and we’ll train a model that specifically recognizes your document type. Think of order references, article numbers, or industry-specific codes.
Is there an on-premise option?
Yes, for Enterprise customers we offer on-premise deployment via Docker containers or as a dedicated server. This is ideal for organizations with strict compliance requirements or very high volumes. Contact us for the possibilities.
How fast is the response time?
Average <2 seconds per page for standard OCR. Complex documents with tables or handwriting can take 3-5 seconds. Batch processing handles up to 60 pages per minute. Need more? Our Enterprise plan provides dedicated resources without wait times.
👨💻 About the Author
Disclaimer: Prices are indicative and dependent on specific requirements. Accuracy percentages are based on tests with high-quality documents.
Actual results vary by document type and quality. For current specifications and SLA terms, please contact us.