Microsoft Foundry: Capacity Issue Preventing OpenAI Model Quota Allocation

Question

Microsoft Foundry: Capacity Issue Preventing OpenAI Model Quota Allocation

alexandru raileanu 0

Problem description

I am experiencing issues with my new Azure AI Foundry subscription that has active Microsoft for Startups sponsorship credits. The subscription remains at 0/0 TPM for OpenAI models like gpt-5.6-sol and terra, and automated quota requests are being denied immediately due to capacity-related messages. I want to understand why the quota requests are denied and if there is a way to move from a Free Tier entitlement to Tier 1 quota.

Environment

Azure AI Foundry / Azure OpenAI quota management, new Microsoft for Startups subscription with active sponsorship credits, quota requests attempted for both Global and Data Local scopes, no specific subscription, tenant, resource, or region details provided.

What I've already tried

I reported the issue around September 17-18, 2026. I used the quota request tool to request additional quota for gpt-5.6-sol and terra. The requests were denied immediately due to capacity issues, leaving the subscription at 0/0 TPM. I reviewed quota guidance indicating that quota is region-based and that sponsorship credits do not automatically grant model quota. I have not tried any other troubleshooting steps beyond these requests.

Current status

The subscription still shows 0/0 TPM for the requested models, with quota requests denied on capacity grounds. I am seeking guidance on whether the subscription can be upgraded from a Free Tier entitlement to Tier 1 quota, and how to resolve capacity limitations preventing deployment.

Sai Kiran Mudavath 185 Reputation points Microsoft External Staff Moderator

2026-09-28T11:02:05.0466667+00:00
Hello @alexandru raileanu ,

Thank you for the detailed information.

Based on the information provided, the behavior does not necessarily indicate an issue with your Microsoft for Startups sponsorship credits. Sponsorship credits provide billing coverage for eligible Azure services, while Azure OpenAI/Foundry model quota is managed separately and is subject to model availability, subscription eligibility, and capacity.

The 0/0 TPM shown for the affected models indicates that no usable quota is currently allocated for that model/scope. Since the quota requests are being rejected immediately with capacity-related messages, we need a few additional details to determine whether this is related to subscription-level eligibility, model availability, regional capacity, or quota allocation.

Could you please provide:

The Azure subscription offer type, such as Sponsorship, Pay-As-You-Go, or MCA.

The Azure AI Foundry resource/project region.

A screenshot of the Quota page showing the affected models.

The exact message returned when the quota request is rejected.

Whether the request was submitted for Global Standard, Data Zone Standard, or a specific region.

Whether other Azure OpenAI/Foundry models also show 0/0 TPM.

Once we have these details, we can determine whether the behavior is related to subscription eligibility, model availability, regional capacity, or quota allocation and advise on the appropriate next step.

Regards,
Sai Kiran
Suhail Ahamed J 0 Reputation points Microsoft External Staff Moderator

2026-10-13T08:57:19.2566667+00:00
Hello @alexandru raileanu ,
I am following up on your post to see if you have had a chance to review our previous response regarding your 0/0 TPM limit and the rejected quota requests.

As your startup works to get these models up and running, could you please share a few additional details so we can best assist you?

The Deployment Region: Which Azure region is your Azure AI Foundry project hosted in?

The Exact Rejection Message: Could you paste the exact error message or code you receive when the automated quota request is rejected?

The Deployment Scope: Are you attempting to request quota for a Global Standard deployment or a local regional deployment?

A Quick Alternative to Try: If you are currently attempting a regional deployment, we highly recommend trying to deploy the model as Global Standard in highly capacitated regions such as Sweden Central or East US. Global Standard utilizes a shared pool of capacity across multiple data centers and often has much higher availability for new or sponsorship subscriptions than regional limits.

If you have already submitted an Azure Support Ticket to have this quota manually adjusted by our engineering team, please feel free to share the Support Ticket ID with us here.

Regards,
Suhail Ahamed J

Your answer

Sai Kiran Mudavath 185 Reputation points Microsoft External Staff Moderator

2026-09-28T11:02:05.0466667+00:00

Hello @alexandru raileanu ,

Thank you for the detailed information.

Based on the information provided, the behavior does not necessarily indicate an issue with your Microsoft for Startups sponsorship credits. Sponsorship credits provide billing coverage for eligible Azure services, while Azure OpenAI/Foundry model quota is managed separately and is subject to model availability, subscription eligibility, and capacity.

The 0/0 TPM shown for the affected models indicates that no usable quota is currently allocated for that model/scope. Since the quota requests are being rejected immediately with capacity-related messages, we need a few additional details to determine whether this is related to subscription-level eligibility, model availability, regional capacity, or quota allocation.

Could you please provide:

The Azure subscription offer type, such as Sponsorship, Pay-As-You-Go, or MCA.

The Azure AI Foundry resource/project region.

A screenshot of the Quota page showing the affected models.

The exact message returned when the quota request is rejected.

Whether the request was submitted for Global Standard, Data Zone Standard, or a specific region.

Whether other Azure OpenAI/Foundry models also show 0/0 TPM.

Once we have these details, we can determine whether the behavior is related to subscription eligibility, model availability, regional capacity, or quota allocation and advise on the appropriate next step.

Regards,
Sai Kiran
Suhail Ahamed J 0 Reputation points Microsoft External Staff Moderator

2026-10-13T08:57:19.2566667+00:00

Hello @alexandru raileanu ,
I am following up on your post to see if you have had a chance to review our previous response regarding your 0/0 TPM limit and the rejected quota requests.

As your startup works to get these models up and running, could you please share a few additional details so we can best assist you?

The Deployment Region: Which Azure region is your Azure AI Foundry project hosted in?

The Exact Rejection Message: Could you paste the exact error message or code you receive when the automated quota request is rejected?

The Deployment Scope: Are you attempting to request quota for a Global Standard deployment or a local regional deployment?

A Quick Alternative to Try: If you are currently attempting a regional deployment, we highly recommend trying to deploy the model as Global Standard in highly capacitated regions such as Sweden Central or East US. Global Standard utilizes a shared pool of capacity across multiple data centers and often has much higher availability for new or sponsorship subscriptions than regional limits.

If you have already submitted an Azure Support Ticket to have this quota manually adjusted by our engineering team, please feel free to share the Support Ticket ID with us here.

Regards,
Suhail Ahamed J