OCR Extension | Version 5

The first and initial version of my OCR extension was released in february this year.
Now it’s time for the first official forum post.

Thanks to ocr.space for this great service.
You can find here more informations about the provider and how to get your own api key for FREE:

Today I have released Version 5 of my extension.

Below you will see what the extension can do for you:

And now a example of how to use all these great properties:

This extension is a full component in makeroid.io
but for every builder that is based on App Inventor available.

This extension works only as apk and not in companion or else.

You like what you see? Now download it and test it by yourself.

Download Ocr Extension
Download Ocr example apk
Download Ocr example aia


thank you for your contribution!

is there any reason for this?

1 Like

It looks like that the companion can not handle multipart http requests.
I have tried it with companion and with usb connection.
But booth without success.
That is the reason for the note that it is only working as apk.

Great Work!
but I think it would perfect if you could apply the automatic resizing of the image to 1MB using your image editing extension, and remove the “\r\n” html line breaks…

You can do the “\r\n” replace by yourself.
Like this:

Maybe I will add a image resizing option in the next version.

Hi. I can´t compile the AIA…I get an error message. Any help?

Mika, please help me. I have this response: “No peer certificate” , what it means?


When I put the OCR and Image Recognizer in the same project, the compiler will have error message:

________Invoking DX
YAIL compiler - DX execution failed.

1 Like

Block ReturnOnlyMessage is not working in version 5

I have the same problem “No peer certificate”. Please help

I am unable to download this extension


first of all thanks for your wonderful extension that I use to parse numbers off a webpage.
Unfortunately sometimes the OCR engine doesn’t recognize the numbers correctly.
I opened a ticket at the forum of ocr.space and nearly immediately got the answer back:
“OCR engine 2 is much better suited for numbers. So specify engine 2 in your API call” (I supplied some JPGs that were used for testing at ocr.space)

And here is the problem:
your wonderful extension doesn’t have an option to specify the engine (1 oder 2).

Would it be possible to get an updated version that has another option “OCREngine”?

Would be very helpful.

Thanks for your efforts.

Best regards, Man

or if that is easier to do:
just supply a hardcoded version of the extension that uses OCR_Engine2.

Thx, Man

Hello Mika,
Thanks for the extension.
Is it normal that it don’t work in AI companion ?
Best regards