Cannot Deploy GPT Models in Azure AI Foundry Using Azure for Students Subscription

Vaibhavi Vaidya 0 Reputation points

I am using an Azure for Students subscription.

I successfully created an Azure AI Foundry project and verified that:

  • Microsoft.CognitiveServices is registered

Microsoft.MachineLearningServices is registered

Azure OpenAI S0 quota is available

The AI Foundry resource and project are both located in East Asia

However, I am unable to deploy any Azure OpenAI model, including gpt-4o, gpt-4.1-mini, gpt-5.1, gpt-5.5, and model-router.

The deployment consistently fails with errors such as:

"No available capacity found for region eastasia with GlobalStandard sku"

"'model-name' with Global Standard SKU isn't available to deploy in your project's region (eastasia)"

I also cannot create Foundry resources in regions such as East US or Sweden Central because of subscription policy restrictions.

Could you please confirm whether Azure OpenAI deployments are currently supported for Azure for Students subscriptions in East Asia, and whether there is a recommended supported region or workaround?I am using an Azure for Students subscription.

I successfully created an Azure AI Foundry project and verified that:

Microsoft.CognitiveServices is registered

Microsoft.MachineLearningServices is registered

Azure OpenAI S0 quota is available

The AI Foundry resource and project are both located in East Asia

However, I am unable to deploy any Azure OpenAI model, including gpt-4o, gpt-4.1-mini, gpt-5.1, gpt-5.5, and model-router.

The deployment consistently fails with errors such as:

"No available capacity found for region eastasia with GlobalStandard sku"

"'model-name' with Global Standard SKU isn't available to deploy in your project's region (eastasia)"

I also cannot create Foundry resources in regions such as East US or Sweden Central because of subscription policy restrictions.

Could you please confirm whether Azure OpenAI deployments are currently supported for Azure for Students subscriptions in East Asia, and whether there is a recommended supported region or workaround?

0 comments No comments
Sign in to comment

2 answers

  1. SRILAKSHMI C 19,110 Reputation points • Microsoft External Staff • Moderator

    Hello @Vaibhavi Vaidya

    Thank you for reaching out to Microsoft Q&A.

    Based on the symptoms you’ve described 'No available capacity found for region eastasia with GlobalStandard sku', “‘model-name’ with Global Standard SKU isn’t available to deploy in your project’s region (eastasia)”

    this behavior typically indicates a regional availability and capacity limitation for the selected deployment type (Global Standard) rather than an issue with your Azure AI Foundry setup or configuration.

    From your description, the following points are confirmed:

    • Azure AI Foundry project is successfully created in East Asia
    • Required resource providers are registered
    • Azure OpenAI S0 quota is available
    • Deployment failures occur across multiple GPT models when using Global Standard SKU

    In Azure OpenAI, model deployments depend on both regional availability and deployment SKU capacity. Even when quota is available, deployments can fail if:

    1. Regional SKU capacity is unavailable

    Global Standard deployments rely on region-specific capacity. In this case, East Asia currently does not have available capacity for Global Standard deployments for the requested model family, which results in the observed error.

    2. Model availability varies by region and SKU

    Not all models (e.g., GPT-4o, GPT-4.1-mini, GPT-5.x, model-router) are uniformly available across all regions or deployment types. Availability is controlled by:

    • Region
    • Model family
    • Deployment SKU (Standard vs Global Standard)

    3. Subscription-level region restrictions (Azure for Students)

    With Azure for Students subscriptions, there may also be policy-driven restrictions on deploying resources in certain regions for example, East US or Sweden Central, which limits the ability to switch to alternative regions with available capacity.

    Please try the following troubleshooting steps:

    1. Verify model availability in the region

    In Azure AI Foundry, check the model deployment dropdown for East Asia to confirm whether the selected model supports Global Standard in that region.

    2. Try alternative deployment SKU

    If supported by the model, switch from Global Standard → Standard deployment

    This may help bypass regional capacity limitations.

    3. Use an alternative supported region

    If allowed by your subscription policy, redeploy the Azure AI Foundry project in a region with higher model availability, such as:

    • East US / East US 2
    • Sweden Central

    4. Validate quota per model and region

    Quota is enforced at a per-model, per-region (TPM-based) level. Even if overall S0 quota is available, the specific model-region combination may still be constrained due to capacity.

    You can review or request adjustments from the Quota management section in Azure AI Foundry.

    Please refer this

    Deploy Microsoft Foundry Models in the Foundry portal (quota + troubleshooting): https://learn.microsoft.com/azure/foundry/foundry-models/how-to/deploy-foundry-models

    Deploy open-source models with managed compute (Preview) (capacity troubleshooting pattern): https://learn.microsoft.com/azure/foundry/how-to/deploy-models-managed

    Create and deploy an Azure OpenAI in Azure AI Foundry Models resource (web-portal) (how to deploy + model availability by region): https://learn.microsoft.com/azure/ai-foundry/openai/how-to/create-resource

    Foundry Models sold by Azure (azure-openai) (example CLI + region availability pointer by category): https://learn.microsoft.com/azure/foundry/foundry-models/concepts/models-sold-directly-by-azure

    I Hope this helps. Do let me know if you have any further queries.

    If this answers your query, please do click Accept Answer and Yes for was this answer helpful.

    Thank you!

    1. Vaibhavi Vaidya 0 Reputation points

      Thank you for your response.

      I have already completed the troubleshooting steps you mentioned:

      • Verified that both the Azure AI Foundry project and Azure OpenAI resource are located in East Asia.

      Registered Microsoft.CognitiveServices and Microsoft.MachineLearningServices.

      Verified that Azure OpenAI S0 quota is available.

      Tested multiple models including GPT-4o, GPT-4.1-mini, GPT-5.1, GPT-5.5, and model-router.

      Checked the deployment wizard and found that Standard deployment is not available. Only Global deployment options are presented.

      Attempted to create resources in recommended regions such as East US and East US 2, but my Azure for Students subscription is restricted by policy and does not allow resource creation in those regions.

      Given that:

      GPT model deployments are failing in East Asia due to capacity/availability errors.

      The subscription policy prevents me from creating resources in the recommended supported regions.

      Could you please confirm whether Azure OpenAI GPT model deployments are currently supported for Azure for Students subscriptions in East Asia?

      If not, is there any supported region available for Azure for Students users, or is a different subscription required to complete Azure AI Foundry model deployments?

      Thank you.

    2. SRILAKSHMI C 19,110 Reputation points • Microsoft External Staff • Moderator

      Hello Vaibhavi Vaidya,

      Thank you for providing the additional details and for confirming that you have already completed the recommended troubleshooting steps.

      Based on the information you shared, we understand that:

      • Your Azure AI Foundry project and Azure OpenAI resource are both located in East Asia.

      The required resource providers (Microsoft.CognitiveServices and Microsoft.MachineLearningServices) are registered.

      Azure OpenAI S0 quota is available.

      You have attempted deployments for multiple models, including GPT-4o, GPT-4.1-mini, GPT-5.1, GPT-5.5, and model-router.

      The deployment experience only presents Global deployment options, and Standard deployments are not available.

      Attempts to create resources in commonly recommended regions such as East US, East US 2, and Sweden Central are blocked by policy restrictions associated with your Azure for Students subscription.

      All deployment attempts in East Asia consistently fail with capacity or availability errors.

      Given these findings, there does not appear to be any misconfiguration in your Azure AI Foundry setup. Instead, this points toward a combination of two factors:

      Regional capacity/model availability limitations in East Asia

      The error messages:

      "No available capacity found for region eastasia with GlobalStandard sku"

      and

      "'model-name' with Global Standard SKU isn't available to deploy in your project's region (eastasia)"

      indicate that the requested Azure OpenAI models are currently not deployable in East Asia using the Global deployment option available to your subscription.

      Subscription-level regional restrictions

      Your Azure for Students subscription is preventing you from creating Azure AI Foundry resources in alternative regions where these models are commonly available. As a result, the typical workaround of selecting another supported region cannot be used in your case.

      Is Azure OpenAI GPT deployment supported for Azure for Students in East Asia?

      Based on the behavior observed, we are currently unable to confirm that Azure OpenAI GPT model deployments are supported for Azure for Students subscriptions in East Asia.

      At this time, there is no publicly documented list identifying which Azure OpenAI models and deployment types are guaranteed to work specifically with Azure for Students subscriptions in each region.

      However, your experience suggests that:

      • Although Azure AI Foundry projects can be created successfully in East Asia using Azure for Students subscriptions,
      • GPT model deployment through the available Global deployment option is either currently unsupported for that subscription and region combination, or temporarily unavailable due to regional capacity constraints.

      Is there a supported workaround?

      Given the restrictions you've encountered, the practical options are:

      Use a different subscription type, such as a Pay-As-You-Go subscription, which typically provides access to a broader set of deployment regions and may allow you to deploy the required models in supported regions.

      Use a shared environment, if available through your institution or instructor. For example, an instructor with access to a supported subscription and region could create the Azure AI Foundry resources and grant students access to work within the existing project.

      Based on the information provided so far, this does not appear to be an issue with resource provider registration, quota configuration, or Azure AI Foundry project setup. Rather, it appears to be a limitation related to the combination of:

      Azure for Students subscription policies, and

      Azure OpenAI model availability/capacity in East Asia.

      Thank you!

    3. SRILAKSHMI C 19,110 Reputation points • Microsoft External Staff • Moderator

      Hi Vaibhavi Vaidya,

      Following up to see if the above answer was helpful. If this answers your query, please do click Accept Answer and Yes for was this answer helpful. And, if you have any further query do let us know.

      Thank you!

    Sign in to comment

  2. AI answer

    Azure OpenAI deployments in Azure AI Foundry are region- and model-dependent and require both:

    1. The model to be available in the target region for the chosen deployment type (Standard/Global Standard/Provisioned), and
    2. Sufficient quota and backend capacity in that region.

    From the provided region-availability data, East Asia is not listed as a supported region for the GPT-4.x / GPT-4o / GPT-5.x families for Foundry Models sold directly by Azure. The available-region tables only include Americas regions such as eastus, eastus2, centralus, westus, etc., for both Standard/Regional and Global Provisioned deployments.

    Because East Asia is not in the supported-region lists for these models, deployments targeting East Asia will fail with errors like:

    • “No available capacity found for region eastasia with GlobalStandard sku”
    • “'<model-name>' with Global Standard SKU isn't available to deploy in your project's region (eastasia)”

    This behavior is consistent with the documented guidance:

    • If a model or version is not available in a region, the recommendation is to select a different region that supports that model and version.
    • For Standard/Regional deployments, prompts and responses are processed only in the region where the deployment exists; if that region does not support the model, deployment is not possible.

    Given the constraints:

    • Azure for Students subscription
    • Policy restrictions preventing creation of resources in supported regions such as East US or Sweden Central

    there is no documented workaround that would allow deploying GPT-4o, GPT-4.1-mini, GPT-5.1, GPT-5.5, or model-router into East Asia, because the models are simply not offered there in the current region-availability matrix.

    The only documented paths to a successful deployment are:

    1. Use a region where the target model is available (for example, eastus, eastus2, centralus, westus, westus3 for GPT-4o/4.1/5.x families) and ensure quota is available.
    2. If using provisioned throughput, ensure PTU quota is available in a supported region and create the deployment there.

    If subscription policies prevent using any of the supported regions, the effective result is that Azure OpenAI model deployments (for these GPT models) are not currently usable with that Azure for Students subscription configuration. In that case, the next step is to work with the subscription administrator or support to adjust policy/region restrictions so that at least one supported region (such as eastus or westus) is allowed.

    There is no documentation in the provided context that indicates a special or different support path for Azure for Students in East Asia, nor any alternative region within East Asia that supports these GPT models for Foundry Models.

    References:

    AI-generated content may be incorrect. Read our transparency notes for more information.
Sign in to answer

Your answer