En | Fr | De | Nl | Es | It

An Innovative Scanning Experience for the Mac.

Prizmo could definitely act as my only scanner. But, it really isn’t only a scanner, it can be used on anything. If you see a movie poster or an ad that you like, and would like to have documented, just take a picture, and Prizmo will help you with perspective, cropping and everything else.
Joe Turner from MacApper.com

Universal Scanning Experience.

Prizmo is a scanning application with Optical Character Recognition (OCR) in 28 languages with powerful editing capability, text-to-speech, iCloud support, and translation into 59 languages. Prizmo can be used by everyone, thanks to its support of VoiceOver.

No matter what capture device's at hand, Prizmo is the key for scanning and recognizing text. It works with pictures taken with your iPhone, iPad, or digital camera, with documents coming from connected or Wi-Fi scanners, even with screenshots! It will help you easily scan any documents (invoices, receipts, boards, giant advertisements...), as well as process business cards.

Prizmo 3.5 introduces new machine learning capabilities including a new built-in OCR and a text translation feature based on deep neural networks. Other modern features include latest macOS enhancements like Handoff, extensions, and third-party exports.
Besides, the user interface has been revamped in order to take advantage of macOS's translucency, vibrancy, and clarity.

More information about these new features.

Optical Character Recognition.

A Powerful Mac OCR Application for Everyone. At last.

Time has come for Mac users who were waiting for a powerful OCR application. Prizmo is a user-friendly and reliable solution all Mac users can afford. It integrates state-of-the-art OCR technologies, and supports multipage documents.

See full list of supported languages below.

Document-specific Processing.

Prizmo includes a specific processing depending on the type of document you want to process:

  • Text: choose the Text mode for processing OCR on regular texts like letters, newspaper articles, magazines, etc. The recognized text can be shared, translated, or read using Text-to-speech technology.
  • Business Card: this mode is suitable for recognizing contact information, and adding it to the address book, or sharing it as a vCard.
  • Image: choose this mode if you want to straighten pictures for instance. No OCR processing will be applied to documents in this mode, only the image processing settings are available for fine-tuning.

OCR Editing and Correction.

Prizmo's built-in OCR Engine is state-of-the-art technology, but it can sometimes be mistaken. That's why Prizmo enables you to update recognized information in several smart ways.

  • Add a region: if a paragraph was not recognized as such, you can add it on the document, and Prizmo will automatically process OCR on the newly added region.

  • Delete a region: imagine Prizmo recognized an image, but you are interested in outputting text only. You can delete unwanted regions with the following button.

  • Merge regions: if Prizmo splits a paragraph in several text regions, select these regions, and merge them in a single one. It will be much easier to proofread this way.

  • Reorder regions: region order matters when outputting recognized information, that's why all regions are numbered (numbers are visible both on the document preview, and in the right column). If the default order is not correct, you can reorder manually.

Besides, you can resize any regions at will.

Recognized information is available in the right column of Prizmo's main window. You can check if everything is correct, but you can also edit text, and correct any mistakes you might find in the text before export. No more post-processing needed!


Prizmo 4
Prizmo 4 - Pro Pack
OCR in 28 Languages Yes Yes
Translation into 59 languages Yes Yes
Neural network-based translation in 22 languages No Yes
Multipage Processing Yes Yes
Perspective Correction Yes Yes
Page Curvature Correction Yes Yes
Text-to-Speech Yes Yes
Batch Processing No Yes
Automator Action No Yes
Custom Export Scripts No Yes
Price $49.99 $74.99
Sold on the Mac App Store Yes
(no trial version available)
Yes, through In-App Purchase
(no trial version available)
Sold through our Store Yes (trial version available) Yes (no trial version available)

Click here for more information about Prizmo 4 incl. the Pro Pack.

Image Correction.

Perspective Correction.

When you take a picture of a document, lighting and centering are seldom perfect. A work environment or a house is indeed very different from a photo studio. Such pictures should be of poorer quality than the images processed with a scanner. It doesn't really matter however as Prizmo is able to correct perspective.

Select the area you want to process on the input photograph. Prizmo will use the grid to straighten the image. From now on, you don't need to take perfect pictures, Prizmo is able to process all images.

Example of how Prizmo corrects perspective
Focus: The image selected in the input picture (left),
and the image Prizmo outputs after correcting perspective and some image processing (right).

Page Curvature Correction.

Page curvature correction window in Prizmo
Have you ever tried to perform OCR on pages of a magazine or a dictionary? Words near the binding are warped, and lines look much more like curves than straight lines. Therefore, words near the binding are usually, and unsurprisingly, pretty badly recognized.

Once again, Prizmo can help you fix this with a unique feature among OCR applications: page curvature correction. This is how curves become straight lines...

It's time for another magic trick. Locate the highest spot of the page on the top side of the grid (and the lowest spot of the page on the bottom side of the grid). Then, use the arrows to imitate the page curvature shape. The image is now ready for performing OCR.

Real example of page curvature correction
Focus: the text on the input photograph (top),
and what the same text looks like after correcting page curvature (bottom).

Lens Distortion Correction.

Calibrating your camera can help you achieve better results when using photographs in Prizmo. With a good calibration model, Prizmo can indeed better correct lens distortion on your photographs.

When importing the input photograph, Prizmo automatically searches for information about used camera model, lens and zoom. You can view the optical data in the Lens window, and there you can also:
  • Select the best calibration model available locally (i.e. among those you previously calibrated).
  • Select the best calibration model among those you previously downloaded from Creaceed's servers.
  • Correct lens distortion manually by moving the radial distortion sliders.
  • Access the camera calibration database, and update the already downloaded calibration models.
  • Create your own calibration model (and decide whether you want to share your calibration model with other Prizmo users or not).

Online Service Integration.

File Sharing through iCloud.

Apple's iCloud makes all your changes everywhere. Your Prizmo documents will appear on your device, automatically. Neither syncing, nor file transfer required.

Prizmo supports iCloud for sharing editable documents across all your Macs, iPhones and iPads. So, you can start processing a very long document with Prizmo on the Mac at the office, and finish it when you are back home on your iPad.

Easily Access your Documents from Anywhere.

Processing OCR with Prizmo is a very good start, but sharing recognized information is usually the next step to be done. That's why Prizmo's output options include:
  • Evernote, for archiving any notes, and for syncing them across your computers and mobile devices,
  • Box, for storing all of your content online, so you can access, manage and share it from anywhere,
  • Dropbox, the online backup, file sync and sharing service,
  • Google Drive, for storing your files, sharing them with exactly who you want and edit them together, from any device,
  • WebDAV, standing for "Web-based Distributed Authoring and Versioning", is a set of extensions to the HTTP protocol which allows users to collaboratively edit and manage files on remote web servers.

Text-to-Speech (TTS).

Listen to Texts at your Pace.

After processing OCR on a document, you can ask Prizmo to read the text out loud for you. You can choose the voice and the speaking rate.

Besides, you have access to navigation settings for playing and pausing speech, and for going straight to the next or previous text region.

Interactive Reading Progress Display.

As soon as speech synthesis gets activated, each word is highlighted in yellow as it is pronounced.

In Prizmo's main window, you can easily recognize the text region which is being read by its blue dotted frame.

High-quality Voices.

Over 90 male and female voices in 26 languages (and their local variants) are available as free downloads. You can download these voices through System Preferences on your Mac. Choose Speech, and click "Customize..." in the System Voice menu to access the whole list of high-quality voices available.

Following languages are supported: Czech, Danish, Dutch, English, Finnish, French, German, Hungarian, Indonesian, Italian, Korean, Norwegian, Polish, Portuguese, Romanian, Russian, Slovak, Spanish, Swedish, Turkish...

Additional Information.

Supported Languages.

Neural Network OCR Languages (28): Arabic, Chinese (Simplified), Chinese (Traditional), Czech, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hungarian, Italian, Japanese, Korean, Norwegian Bokmål, Polish, Portuguese, Romanian, Russian, Serbian (Cyrillic), Serbian (Latin), Slovak, Spanish, Swedish, Turkish, Ukrainian.

Translation Languages (59): Afrikaans, Arabic, Bangla, Bosnian, Bulgarian, Cantonese (Traditional), Catalan, Chinese (Simplified), Chinese (Traditional), Croatian, Czech, Danish, Dutch, English, Estonian, Fijian, Filipino, Finnish, French, German, Greek, Haitian Creole, Hebrew, Hindi, Hmong, Hungarian, Indonesian, Italian, Japanese, Klingon, Korean, Latvian, Lithuanian, Malagasy, Malay, Maltese, Norwegian Bokmål, Persian, Polish, Portuguese, Romanian, Russian, Samoan, Serbian (Cyrillic), Serbian (Latin), Slovak, Slovenian, Spanish, Swahili, Swedish, Tahitian, Tamil, Thai, Tongan, Turkish, Ukrainian, Urdu, Vietnamese, Welsh.

Neural Network Translation Languages (22): Arabic, Bulgarian, Chinese, Chinese, Czech, Danish, Dutch, English, French, German, Hindi, Italian, Japanese, Korean, Norwegian Bokmål, Polish, Portuguese, Romanian, Russian, Spanish, Swedish, Turkish.

An example text for certain languages is available here.