scanmydocz This tool uploads your file Scan a document

Tools / Image to Text

Image to text

Real recognition. Reads a photograph, not just a text layer.

Image to Text runs on a paid model, so it is part of Pro rather than the free tier. Every other tool here is unaffected. See the plans.

Choose a JPEG, PNG or WebP

or drop it here

How it works

  1. Add a photograph or a scan.
  2. Choose what kind of page it is, if it is handwriting or a receipt.
  3. Read the text, then copy it or download the .txt file.

Reading words off a photograph

A photograph of a page contains no text, only pixels arranged to look like text, which is why extraction returns nothing and recognition is a different job.

Light matters more than megapixels

Even, indirect light beats a high resolution photograph with a shadow across half the page. Turn the flash off and move to a window.

Get the page square in the frame

Recognition reads lines. A page photographed at an angle gives it curved lines to follow and accuracy falls.

Tell it what kind of page it is

Handwriting, a receipt and a printed page are read by different models. Choosing the right one is usually a larger gain than retaking the photograph.

Read the output before you trust a number

Recognition is confident and occasionally wrong. A 5 read as an S in an account number is the kind of error that costs something.

Where this is the wrong tool. This one uploads. The image is sent over an encrypted connection, read on our servers and deleted after 24 hours, because recognition of this quality does not run in a browser tab. If the document is sensitive, that is the trade you are making.

When the browser is not enough

Scanly for Android

Take document-scanning workflows to your Android phone with Scanly. See the Google Play listing for current features, requirements and data-processing information.

Get Scanly for Android

scanmydocz Pro

See the current plan features and limits before subscribing.

See plans

Questions

Is this different from PDF OCR on this site?

Yes, and the difference matters. PDF OCR pulls out a text layer that is already in the file, and tells you plainly when there is not one. This reads the picture itself. If your PDF is a scan, this is the tool that can help and that one cannot.

Does my image leave my device?

Yes. Recognition of this kind cannot run in a browser tab at usable quality, so the image is sent to our servers, read, and deleted 24 hours later. That is why this tool carries a different badge from the ones that finish on your own machine.

Is there a limit?

Without an account, six pages an hour and 25 MB a file. That is a cost limit rather than an arbitrary one: every page costs us money to read.

Will it read handwriting?

Often, and not always. Choose the handwriting option and it will use a model better suited to it. Neat writing on clean paper does well; a doctor's note on a crease does not, and no tool anywhere will tell you otherwise honestly.

Tools people reach for next to this one