How recognition works
Text recognition (OCR) needs a lot of computing power and large models, so Image Text to Text does it on our server instead of in your browser. When you capture an area or upload an image, the extension sends that image to the server, the server returns the text, and the extension shows it to you.
What is sent and why
| Data | Why |
|---|---|
| The image or screen area you choose | To recognise the text in it |
| A random device ID | To count free recognitions and check your plan |
| The e-mail you enter when you buy | To link your plan |
What is not collected
- your browsing history or the pages you visit;
- screenshots you do not choose to recognise;
- the text after it has been returned to you.
Good practice
Do not recognise images with passwords, payment card numbers or documents you are not allowed to share. For text that is already text on a page, copy it directly or use Highlight and Copy Protected Text — nothing is sent anywhere.
The full policy is on the privacy page.