Umi-OCR Offline OCR Software for Windows & Linux
Umi-OCR turns screenshots, images, and PDFs into text with 100% offline recognition - no uploads, no data leakage.
41K+
GitHub Stars
100+
Languages
100%
Offline
Powerful Features
Everything you need for text recognition
Screenshot OCR
Instantly capture and recognize text from any area of your screen with a simple hotkey
Batch Processing
Process hundreds of images at once with high accuracy and no quantity limits
PDF Scanning
Extract text from scanned PDFs and create searchable double-layer PDF documents
QR Code Recognition
Scan and decode QR codes and 19 types of barcodes from images
Math Formula
Convert mathematical formulas in images to LaTeX format
100% Offline
All processing happens locally on your computer - your data never leaves your device
Frequently Asked Questions
Yes, Umi-OCR is completely free and open-source under the MIT license. There are no hidden fees, subscriptions, usage limits, or watermarks.
No, Umi-OCR works entirely offline. All text recognition is performed locally on your computer, ensuring your documents and images remain private.
Paddle version offers higher accuracy with a larger model (~128MB), ideal for complex documents. Rapid version provides faster processing with a smaller footprint (~97MB), perfect for quick tasks.
Umi-OCR supports over 100 languages for text recognition, including English, Chinese, Japanese, Korean, Russian, German, French, and most European languages.
Yes, Umi-OCR can recognize both printed and handwritten text, though accuracy may vary based on handwriting clarity and style.
Umi-OCR runs on Windows 7 or later and Linux x64 (Debian). It requires approximately 200MB of disk space and works on both high-end and low-CPU systems.
Download Umi-OCR
Choose your preferred version and start recognizing text today
Latest Version: v2.1.5