OCR Privacy Guide

ScanText is built for privacy-first OCR. By default we process images in your browser when possible. This page explains what happens to your files in each mode.

Many free OCR sites upload every file to the cloud by default. ScanText reverses that model: browser OCR runs locally with Tesseract when your device supports it, which means photos of IDs, invoices, or personal notes can stay on your machine for that step.

When you enable clearer scan or when browser confidence is low, a single image may be sent to our API on Railway over HTTPS. We do not require an account, we do not use your files for training, and temporary uploads are deleted within about 60 seconds after processing.

You can verify behavior yourself: use browser developer tools network tab during a default run — you should see no image upload until clearer scan is triggered. For regulated workflows, combine on-device OCR with your own retention policies and avoid clearer scan for highly sensitive material.