How OCR works
Each PDF page is rendered to an image in your browser and passed to the OCR engine. The first run may take longer while language data is downloaded.
OCR is CPU- and memory-intensive. For long scanned documents, process up to about 50 pages at a time.
Privacy
In the current public mode, OCR runs directly in your browser: only the recognition engine and language data are loaded, while the document itself is not uploaded to our server. Server mode is temporarily disabled.