Azure Ai Document Intelligence on local .PDF file with powerautomate desktop

Eric Dallaire,

To process a local PDF file with Azure AI Document Intelligence in Power Automate Desktop, you need to ensure the file is correctly passed as binary data. The error suggests that the system expects either a binary file or a Base64-encoded string.

Steps to Fix the Issue:

Convert the Local PDF File to Binary Data:

In Power Automate Desktop, use the "Read File" action to read the local PDF file (C:\Temp\Test.pdf) as binary data.

Store it in a variable, e.g., %BinaryData%.

Pass Binary Data in the API Call:

Instead of passing the file path (C:\Temp\Test.pdf), pass the binary content.

Use the %BinaryData% variable in the Document/Image File Content field (as shown in the second screenshot).

Use the Correct API Parameter (base64Source):

If the API still requires a Base64-encoded string, you must convert the binary data to Base64 before passing it.

Use the "Convert File to Base64" action in Power Automate Desktop.

Then, pass the Base64 string as the base64Source parameter.

Hope this helps, do let me know if you have further queries.

Thanks.

If i convert my local file in binary like this

And use it like this

I receive this error during runtime

ID de corrélation : ea5a23e3-e777-4afa-9ded-b92d03ec0a0f

Informations supplémentaires :

"StatusCode": "400",

"Message": "",

"Body": {

*"error": {*

  *"code": "InvalidArgument",*

  *"message": "Invalid argument.",*

  *"innererror": {*

    *"code": "ParameterMissing",*

    *"message": "The parameter urlSource or base64Source is required."*

I've tried converting the file in base64 but it say Document/Image File parameter must be "Binary"

There is no place to fill the base64Source parameter in PowerAutomate