Paul Digital
Menu

JPG to Text

Read the text out of a JPG photo or scan using OCR that runs on your device.

Drag and drop a JPG / JPEG image here

Accepts JPG / JPEG · one image · up to 50 MB · English text

Text is recognised on this device. Your image is never uploaded.

English printed text only. Accuracy depends on image quality, and handwriting, tiny or stylised text and low contrast are often misread. Always proofread the result.

How to use JPG to Text

  1. Choose a JPG / JPEG image, or drag it onto the drop area. The text should be clear and in English.
  2. Press Extract text. The first run downloads the OCR engine (about 8 MB) from this site and caches it.
  3. Wait for recognition to finish. A progress message shows what is happening, and you can cancel.
  4. Proofread the text, check the confidence score, and edit mistakes directly in the box.
  5. Copy the text or download it as a .txt file.

About this tool

This tool uses Tesseract, an open-source optical character recognition engine compiled to WebAssembly, to recognise printed text in your JPG. The image, the engine and the English language data all stay on your device. The OCR files are served from Paul Digital itself, not from a third-party CDN.

OCR is never perfect. Accuracy is best with sharp, well-lit, straight printed text in English at a decent size. It is poor for handwriting, decorative fonts, tiny text, curved pages and low-contrast photos. If nothing is recognised the tool says so instead of showing an empty success. Always proofread the result.

Formats and browser compatibility

Accepted input: JPG / JPEG.

Needs a browser with WebAssembly and Web Workers (all current Chrome, Edge, Firefox and Safari). The first run downloads about 8 MB of OCR engine and language data, which the browser then caches. Large images take longer and use more memory; images over 25 megapixels are rejected.

Only English is supported at present, so other languages and scripts will be recognised poorly or not at all.

Practical uses

  • Copy text from a photo of a printed page or sign.
  • Pull text out of a screenshot saved as JPG.
  • Digitise a receipt or letter for searching.

Tips for better results

  • Photograph text straight-on with even lighting and no glare.
  • Crop to just the text area; extra background can add noise.
  • Use a higher-resolution image; text should be at least about 20 px tall.

Frequently asked questions

Is my image uploaded for OCR?

No. Recognition runs in your browser. The engine and language file are downloaded from this site, but your image never leaves your device.

Why is the text wrong or missing?

OCR struggles with blur, small or stylised text, handwriting and low contrast. Try a sharper, higher-contrast, cropped image.

Does it support other languages?

Not yet. Only English is installed, so other languages will not be read reliably.

Are my files uploaded anywhere?

No. Your image is processed in your browser. The OCR engine and English language file are downloaded from Paul Digital itself, never from a third party, and your image is not sent anywhere.

Related tools

Privacy

Your image is processed in your browser on your own device and is not uploaded. The OCR engine files are served from Paul Digital rather than a third-party network, and nothing is stored once you close the page.