Umi-OCR
FreeScreenshot-to-text entirely on your machine — unlimited, no account
An offline OCR tool. Screenshot recognition, batch image processing, turning scanned PDFs into searchable text and QR code reading all run locally with no image uploaded. It handles Chinese, English, Japanese, Korean and more, including vertical text and formulas, and exposes a command-line and HTTP interface for other programs.
Why we picked it
MIT-licensed, 47,400 stars.
Online OCR sites stack three problems: a daily page limit, a sign-up, and your image going to their server. For contracts, identity documents or internal paperwork, the third is a problem on its own.
Umi-OCR puts the engine on your machine, works offline, and has no limit. The screenshot entry point gets the most use — a shortcut selects any region of the screen and the text lands in the clipboard, pulling words out of pages that block copying, out of PDFs, out of video frames.
Batch mode takes hundreds of images at once, or turns a scanned PDF into a searchable one with a text layer. Only after that step can Stirling PDF or pdfsam, both listed here, do anything with the content.
To be clear about:
- Windows and Linux only; there is no macOS build.
- The most recent release was November 2025. It is complete and usable, but the pace has slowed.
- Accuracy on printed text is strong; handwriting and blurry photographs drop off noticeably.