← Back to Discover

Umi-OCR

Free, open-source desktop OCR program for Windows and Linux that reads text from screenshots, image batches and PDFs without a network connection, and also scans and generates QR codes.

Docs & workspaces
Visit website ↗

What it does

Umi-OCR is a text recognition program that runs entirely on your own computer, with no network needed. You unpack the release archive and launch Umi-OCR.exe, with no installer step. The project is open source under the MIT license and describes itself as completely free. It is organized into tabs. Screenshot OCR recognizes text from a captured region or a pasted image, batch OCR works through image files with no limit on how many you import, and document recognition handles PDF, XPS, EPUB, MOBI, FB2 and CBZ files, including output as a searchable PDF. Batch results can be saved as txt, jsonl, md or csv. Text post-processing tidies the layout and reading order of what was recognized, and an ignore-area feature lets you draw boxes so watermarks, headers or footers are left out. The tool also scans and generates QR codes and barcodes, supports several interface languages, and can be driven from a command line or an HTTP interface. Different offline OCR engines can be swapped in through plugins.

Capabilities

  • Offline text recognition with no network required
  • Screenshot OCR from a hotkey or a pasted image
  • Batch OCR of image files with ignore areas for watermarks
  • PDF and ebook recognition with searchable PDF output
  • QR code and barcode scanning and generation
  • Command-line and HTTP interfaces

When to use it

  • You want to pull text out of screenshots or images without sending them to an online service.
  • You have hundreds of scanned pages or images to convert in a single run.
  • You want to call OCR from your own scripts through a command line or HTTP.