Azure OpenAI quota does not decrease even if the model is deleted
Hello,
I'm currently working on a proof of concept using Azure OpenAI. I previously deployed GPT models (like GPT-4 and GPT-4o) using the Standard deployment type. In order to free up quota for new deployments, I have deleted all previously provisioned models from my Azure OpenAI resource.
However, even after deleting all deployments, the quota still appears to be in use, and I am unable to create any new Standard deployments. I have waited over 24 hours and verified that no active models remain in the resource.
My questions are:
- Is there any delay or hidden allocation mechanism that might be causing the quota to remain blocked?
Will the quota be automatically released after a certain period?
- Is there any recommended way to force quota refresh or manually request its release (other than creating a support ticket)?
Here’s what I’m using in the PoC:
Model: GPT-4o (preferred), GPT-4
Embedding model: text-embedding-3-small
Tools: Azure OpenAI, Azure AI Search, Azure AI Foundry, Azure Blob Storage
Thank you in advance!