@Citizen Dev I believe you are referring to the batch transcription API of the speech service.
Are you referring to pay as you go account of your Azure resource or a standard tier resource of the endpoint that is being used?
If you are using a paid subscription of Azure but your resource is created with free tier then the batch transcription API will not work since standard tier of the speech resource is required for batch transcription API. You can refer the link mentioned above for more details.
Are you using a shared or common cognitive service resource or a dedicated speech service resource?
If you are using a shared cognitive service resource then I would recommend to use the dedicated speech resource with standard tier S0 and then check the same with the new endpoint.
I hope the above suggestions help.
If an answer is helpful, please click on or upvote which might help other community members reading this thread.