Umi-OCR

Use it when

• Hundreds of local images or screenshots need their text extracted in batch, fully offline
• A scanned PDF must become a searchable dual-layer PDF without uploading the document to any cloud service

What it solves

Not the fit when

• macOS (Windows 7 x64 and Linux x64 only)
• GPU-accelerated or online OCR engines (planned, not shipped)
• image translation or fixed-region monitoring (future roadmap items)
• speech or audio transcription
• table extraction to Excel (listed as a future plan only)
• hosted cloud OCR API at service scale

Install

Download the .7z or self-extracting release from GitHub releases and extract (no install needed; run Umi-OCR.exe on Windows or umi-ocr.sh on Linux), or on Windows: scoop bucket add extras && scoop install extras/umi-ocr

Invoke

Launch Umi-OCR.exe; use the screenshot OCR hotkey, the Batch OCR tab for folders of images (txt/jsonl/md/csv output), or the Document tab to turn scanned PDFs into dual-layer searchable PDFs; also callable via documented CLI and HTTP interfaces

Alternatives

No reviewed alternatives recorded yet.