Umi-OCR is a free, fully offline OCR application with 48,000+ stars on GitHub that turns screenshots, images and PDFs into editable text entirely on your own computer.
Cloud OCR services are slow, often paid, and a privacy problem for sensitive documents. Umi-OCR takes the opposite approach: everything runs locally, powered by PaddleOCR recognition engines, wrapped in a fast Qt/QML desktop interface. Press a hotkey, select any region of the screen — an image, a video frame, a scanned page — and the recognized text lands in your clipboard. For larger jobs it batch-processes folders of images and extracts text from PDF documents, with layout options to drop headers, footers and watermarks and to merge paragraphs sensibly. It also scans and generates QR codes, recognizes many languages with built-in language libraries, and exposes command-line and local HTTP interfaces so scripts and other tools can call it. Released under the MIT license by hiroi-sora, it is one of the most popular open-source OCR applications in the world.
Umi-OCR is completely free and open source under the MIT license. There is no premium version and no usage limit — download it and recognize as much text as you like.
Office workers, students, translators and researchers who regularly pull text out of images, screenshots and scans — especially where documents are confidential or the internet is not an option.