OCR (Scanned → Text) — free, private how-to guide
Run OCR on scanned PDFs and images in your browser to recognize Georgian, English, and Russian text. Turn a scan into selectable text or a searchable PDF — privately, with nothing uploaded. The Georgian language models make it ideal for local legal documents.
Updated
Step by step
Select a file
Drag and drop a scanned PDF or a JPG/PNG image.
Pick the languages
Choose every language that appears in the document for accurate recognition.
Copy or download
Get the recognized text to copy or save as a .txt file.
Open OCR (Scanned → Text)Common use cases
- Make a scanned Georgian contract searchable
- Copy text out of a photographed document
- Build a searchable archive of scans
Frequently asked questions
Does my document get uploaded for OCR?
No. Recognition runs in your browser via WebAssembly. The only thing fetched from the network is the OCR engine and language model, which download once and are cached — your document itself never leaves your device.
Which languages are supported?
Georgian, English, and Russian, individually or combined. Select all the languages present in the document so mixed-language pages are read correctly.
Why is OCR slower than the other tools?
OCR is computationally heavy — each page is analyzed pixel by pixel. Larger or multi-page documents take longer, and the first run also downloads the language model.
Does it really recognize Georgian?
Yes — Georgian, English, and Russian are supported, including mixed-language documents.
Can it create a searchable PDF?
Yes — choose “Searchable PDF” to keep the original look with an invisible, selectable text layer.