Skip to main content

The Whisper-1 model is showing in the Zapier OpenAI action which is great.

I have an incoming audio from Airtable that is available as an attachment or URL link so should work.

However when you select the Whisper-1 model it only shows Prompt. 

But for the Whisper-1 model the required params are file and model. 

Prompt, langauge and response format are optional.

Anyone get this working?

Hi @Reilly 

Welcome to the Community!

I noticed that you’re already working with one of our Support Specialists on this. They’re asking,

Can you explain the issue in a little more detail? Is there a specific value you were hoping would show up in the trigger step but isn't showing?

If you need further assistance on this, you may continue working with our Support Specialist as they could investigate more.

Thanks! 😊


Is there any update on this?

 

 


Hi @CameronTheAiGuy 

Do you also have issues with OpenAI in your Zap? If so, what kind of error message are you receiving?


@Brem Anyone using the Whisper-1 model would. The model requires file and model as required params, everything else is optional. The Zapier OpenAI action only provides prompt.


If trying to call the Whisper-1 model via Post in Webhooks by Zapier and passing the params as multi-part/form-data we keep getting an error that the model is not passed, even though it is. note: Whisper-1 requires the file to be included in the call. Have this working in PHP if anyone wants it.

 


Hi @Reilly 

I see that you’ve been working with our Support team and you’re in conversation with them. It will be best to keep the conversation going over with them because they can dig deep into the error logs of your Zap to look for the possible cause of errors.


Did you get this working seams passing attachement from email not working?

 


Hi @cksolutions,

Just giving a quick update here, it seems like OP was running to feature where the File field is not available when using the whisper-1 model in Send Prompt action. That said, this issue has now been added to the whisper-1 model and should be able to send files now.

Hopefully, this helps! 😊