OCR Scanner
ImageExtract text from images and scans with on-device OCR.
OCR Scanner pulls the text out of a screenshot, a photo of a page, or a scanned PDF and gives you something you can actually copy. Pick the language, optionally sharpen a low-contrast scan, and the recognition runs on-device using a local OCR engine — no upload step, which matters when the thing you're reading is an invoice, an ID or a contract. Copy the result or download it as a .txt.
Why use OCR Scanner
100% private
Your image files are processed locally and never uploaded.
Instant
No upload or download wait — it runs at the speed of your device.
Free to start
Use it free every day; go Pro for unlimited runs.
No install
Works in any modern browser. No sign-up to begin.
How to use OCR Scanner
- 1Add the image
Drop in a photo of a page, a screenshot or a scan containing text.
- 2Pick the language
Choose English, Norwegian, Swedish, Danish or a combined pair — the right model matters more than anything else for accuracy.
- 3Read and copy the text
The recognised text appears as selectable text you can copy or download; the whole recognition runs on your device.
Frequently asked questions
Which languages can it read?
English, Norwegian, Swedish and Danish, plus combined pairs like Norwegian + English for documents that mix the two. Choosing the right language matters more for accuracy than anything else — an English model reading Norwegian will mangle æ, ø and å.
Why is the recognised text wrong or full of junk?
Almost always the source image. OCR needs sharp, high-contrast, roughly straight text: a blurry or angled photo, a low-resolution screenshot, or pale text on a busy background all produce garbage. Re-take the photo square-on in good light, or crop tight to just the text, and accuracy jumps.
Can it read a PDF?
Not directly — this tool reads images. For a text-based PDF use PDF to Text, which extracts the real text and is exact. For a scanned PDF, export the pages as images with PDF to Image first, then run them through here.
Is my document uploaded to be read?
No. The recognition engine is downloaded to your browser and runs on your device, so a payslip, a passport page or a confidential contract is never sent anywhere — which is unusual for OCR, since most services process on their servers.