How to stop a realtime endpoint VM in Azure AI Studio?

Question

I recently deployed a LLM with Azure AI Studio as a realtime endpoint with shared quota. When i deployed it i selected had to select a VM with a few dollars per hours of cost. However the costs soon exploded even when i was not using the endpoint. I had not used the endpoint for more than 2 hours and the costs far exceeded 2 hours of usage.
I know that you must stop/deallocate VMs if you are not using them in order to lower the costs but i couldn't find the selected VM anywhere. Not in the Azure AI Studio nor in the Azure Portal.
I am by quite new to Azure when it comes to these things, so i would appreciate if someone could explain to me how the costs came to be and how to prevent the costs from accumulating this way. I am also not entirely sure if i understand quotas properly, so an easy explanation in that regard would also be highly appreciated.
Also on a side note: The OpenAI Models don't seem to be in the deployment categories of pay-as-you-go or realtime endpoint. How exactly do the costs for those work?

Answer 1

@a23829499 If you see the compute type is managed, you will not be able to stop the VM behind this endpoint.

You will have to delete the deployment if you no longer need the endpoint. If you believe that the costs have been incurred after deleting the endpoint you can raise a support case for billing enquiry through azure portal. I hope this helps!! Thanks!!

If this answers your query, do click Accept Answer and Yes for was this answer helpful. And, if you have any further query do let us know.