The short version: SnapText does not collect, store, transmit, sell or
share any of your data. All text recognition (OCR) happens locally inside your browser. Your
images and the text extracted from them never leave your computer.
What SnapText processes
Images you provide — when you drag in, paste, pick or capture an image,
SnapText runs OCR on it locally using a bundled engine (Tesseract.js). The image is
processed in memory in your browser and is never uploaded.
A captured screen region — “Capture a screen area” takes a one-time
screenshot of the current tab, crops it to the region you select, and runs OCR on that crop
locally. The crop is stored only briefly in your browser’s local storage to pass it to the
result page, then removed once it has been read.
What SnapText does NOT do
It does not upload your images, screenshots or recognised text.
It does not require an account, login or sign-up.
It does not track you, use analytics, or show advertising.
It does not read or monitor your browsing — it only acts on the current tab at the moment
you start a capture, and only to take that one screenshot.
It does not fetch or run any remote code; the OCR engine and language data are bundled in
the extension.
Permissions
activeTab — to take a one-time screenshot of the current tab when you
start a capture.
scripting — to show the selection overlay on the page only while you are
capturing.
storage — to briefly hold the cropped image locally and to remember
interface preferences.
Data retention
Recognised text stays in the extension window until you close it or start over. The temporary
captured crop is deleted from local storage as soon as the result page reads it. Nothing is
kept on any server, because nothing is sent to one.