Free
Trial
One Connector, Three Engines
Windows.Media.Ocr
Built into Windows 10 and later. No extra install. Supports the most languages.
Apple.Vision
Built into macOS 10.15 and later. No extra install.
Onnx.Ocr
An ONNX Runtime-based engine for Windows 10 and later. Supports page orientation detection.
Choose Your Engine
Switch engines and languages with a single setting. Your scanning and editing workflow stay the same.
35+ Languages
Cover Latin, Cyrillic, Greek, Chinese, Japanese, Korean, Arabic and more across the supported engines. Arabic output is correctly shaped and ordered in PDFs.
Searchable PDF Output
Export plain-text PDFs or add an invisible text layer over the original image. Save as PDF/A-1b, 2b, or 3b for archiving.
Region-Based OCR
Recognize only the areas you select instead of the whole page.
Local Processing, No Cloud Upload
OCR Connector processes images on the user’s device and does not send images or OCR results to Dynamsoft or any remote server. Dynamsoft does not use your documents for AI training.
Digitizing invoices, contracts, and forms for automated processing
Extracting specific sections from scanned documents without capturing unnecessary data
Generating searchable PDFs for archiving and compliance
Multi-language document workflows in global enterprises