Our Models Explained

We offer two powerful OCR models, each tailored for specific needs. Understand the difference to choose the best one for your task.

Model A

Cost/Benefit Model

This model provides an excellent balance of high performance and affordability. It's optimized for common document and image extraction tasks with exceptional accuracy for printed and handwritten text in Latin-based alphabets.

Accuracy

~95-98% for standard digital and print OCR.
~90% for handwritten text (Roman/Latin alphabet).

Best For

Extracting text from invoices and receipts.
Digitizing printed documents and forms.
Processing handwritten notes and letters.

Model B

Wide Languages Model

Our most advanced model, designed for global applications. It leverages a larger language base to understand and extract text from a vast array of languages and character sets, including complex scripts.

Language Support

Extensive coverage of global languages.
Superior handling of non-Latin scripts (e.g., Japanese, Chinese, Korean, Cyrillic).

Best For

International document processing.
Extracting text from media with mixed languages.
Applications requiring the highest level of language flexibility.