0/0 TPM for all GPT-5.6 / GPT-6 models on a new startup subscription: how to get baseline quota?

Question

0/0 TPM for all GPT-5.6 / GPT-6 models on a new startup subscription: how to get baseline quota?

Michael Trifonov 0

We're an early-stage startup on Azure startup credits. Our subscription has no usable quota for any current GPT model, so we can't evaluate them at all.

Quota page (Show all): every GPT-5.x and GPT-6 model shows 0/0 TPM under Global scope (gpt-5.6-sol/luna/terra, gpt-6-sol/luna/astra, gpt-6.1-sol, gpt-5.5, gpt-5.4, etc.). Data Zone Standard is also 0/0.
Deploying these models fails with "Insufficient quota" for both Global Standard and Data Zone Standard.
Instant access lists them, but the playground is disabled: "Insufficient capacity to use this model instantly."
Non-OpenAI Direct-from-Azure models (e.g. grok-4.3) deploy fine with 1M TPM on the same subscription, and gpt-5-mini shows 500K TPM.

Attempted:

Quota increase form for gpt-6-astra: denied within minutes ("region not available to expand capacity"), though the form doesn't offer a region choice for this model.
A second quota form request (gpt-6-sol): pending.
Support tickets: redirected here, since our Developer support plan doesn't cover technical/quota issues.

We want to experiment across the current GPT-5.x / GPT-6 lineup to figure out which models fit our use case. Is there a way to get the subscription moved off this zero-quota state (e.g. to a higher quota tier) so we have a small baseline allocation across these models? If not, what's the recommended path for a startup that needs to evaluate several models? Happy to share subscription details privately.

Sai Kiran Mudavath 175 Reputation points Microsoft External Staff Moderator

2026-10-03T09:21:33.4833333+00:00
Hello @Michael Trifonov ,

The 0/0 TPM values indicate that the subscription currently has no quota available for those specific model and deployment-type combinations. The fact that gpt-5-mini shows 500K TPM and another Foundry model can be deployed does not establish quota availability for the other GPT-5.x and GPT-6 models, because quota is assigned separately by model and applicable deployment scope.

For models including gpt-6-astra, gpt-5.6-sol, gpt-5.6-terra, gpt-5.6-luna, and gpt-5.5, Microsoft documents that no separate model-access request is required, but a quota request may still be required depending on the subscription’s quota tier. Tier 5 and Tier 6 subscriptions receive quota for these models by default.

Microsoft does not document a self-service setting that lets a customer move a subscription directly to a selected higher quota tier or obtain a small baseline allocation across the entire model lineup. Microsoft for Startups credits should therefore not be interpreted as documented confirmation that quota is available for every model.

Since the gpt-6-sol quota request is pending, that request must be evaluated through the quota process. For the gpt-6-astra request, the rejection states “region not available to expand capacity.” The published documentation does not explain why the request form omitted a region selection for that model, so we cannot determine the reason for that form behavior from public documentation alone.

For immediate evaluation, use the models for which the subscription already shows available quota, such as the reported gpt-5-mini allocation. For each additional required model, request quota through the Foundry quota experience. Quota approval and capacity availability cannot be guaranteed.

References:

Manage Azure OpenAI quota in Microsoft Foundry

Azure OpenAI reasoning models and quota requirements

Azure OpenAI quotas and limits

0/0 TPM for all GPT-5.6 / GPT-6 models on a new startup subscription: how to get baseline quota?

Your answer