Privacy.

The short version.

Your images never leave your device. This tool runs optical character recognition entirely in your browser using an open-source on-device OCR engine. We have no server endpoint that receives images, no image storage, and no way to see your files or your extracted text, because they never reach us.

What happens when you use the tool.

1. You select, drop, or paste an image. It is decoded by your own browser.
2. Light preprocessing (resizing, grayscale, contrast) runs on a canvas in your tab.
3. An OCR worker in your tab recognizes the text.
4. The result appears in an editable box on your screen only.
5. Closing or refreshing the page discards everything. Nothing persists unless you download the .txt file yourself.

What we do not collect.

• No image contents: images are never transmitted to us.
• No extracted text: recognition output stays in your tab.
• No file names tied to content: we never see your files at all.
• No accounts: there is nothing to sign up for, so there is no identity to track.

What we may collect.

We keep analytics to the minimum needed to operate the site. If analytics are enabled, they record only anonymous product events such as page views, OCR started, OCR completed, OCR failed, copy clicked, and download clicked. These events contain counts and technical metadata only, never image contents, file names, or extracted text.

Standard, privacy-respecting web-server logs (such as requested URLs and coarse technical headers used for security and reliability) may exist at the hosting layer. They are never combined with document contents, because document contents never reach any server.

Third-party resources.

The page itself, fonts, and the OCR engine files (WebAssembly and language data) are fetched from standard hosting/CDN infrastructure like any website. Fetching these static resources does not transmit your images. Your browser downloads the engine, then runs it locally on pixels that stay put. No OCR API is ever called.

Cookies and local storage.

We do not use tracking cookies. Your browser may cache the OCR engine files so repeat visits are faster. That cache lives on your device and contains no personal content. We do not currently store OCR history; if a local history feature is ever added, it will use your browser's own storage only, stay on your device, and be clearly disclosed here first.

Children.

The tool requires no personal information from anyone, including children. Since images and text never leave the user's device, there is no child data for us to collect, store, or share.

Changes to this policy.

If this policy changes, the updated date at the top will reflect it. Any change that weakens the local-processing guarantee above would be a fundamental product change, not a quiet edit.

Contact.

Questions about privacy? Reach us via the contact page.