m unable to deploy any models in Microsoft Foundry using my subscription.

Hi lin wang,

If you are unable to deploy any models in Microsoft Foundry, this usually happens because of region or project setup, not because something is broken.

Most models can only be deployed from specific regions. If your Foundry project is created in a region where the model is not supported, the portal will show no available regions and block the deployment. In that case, creating a new Foundry project in a supported region usually fixes the issue.

Another common reason is quota. Even if access is approved, some models start with zero quota. When quota is zero, deployment will not work. Once a supported region appears, it is also important to check that quota is available for the selected model.

Some models are also limited to certain deployment types. If the deployment type does not match what the model supports, you will not be able to proceed.

In most cases, once the project region, model region availability, and quota are aligned, model deployment works without any further changes.

Official references that explain this behavior clearly.

Microsoft Foundry model deployment and region availability https://dotnet.territoriali.olinfo.it/azure/ai-studio/how-to/deploy-models

Model regional availability and limitations https://dotnet.territoriali.olinfo.it/en-us/azure/foundry/openai/concepts/model-retirements

Quota management in Azure AI Foundry https://dotnet.territoriali.olinfo.it/azure/ai-studio/how-to/quota

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

Thankyou!

"No available regions" in the deployment dialog means that, for the selected model, there is currently no supported combination of:

• Model availability for the subscription’s billing country/region, and
• A Foundry project (hub) region that is allowed for that model.

For GPT and other Foundry Models, deployment is only possible when all of the following are true:

1. The model is offered in at least one region for the subscription’s billing country/region.
2. There is a Foundry project in one of the model’s supported regions.
3. The model is enabled for the subscription (for gated models) and there is sufficient quota.

If any of these conditions fail, the portal shows No available regions and lists all other regions as Region not supported.

To move forward, use these checks and actions:

1. Verify model regional availability
   • Check the model’s supported regions in the documentation for Region availability for models in standard deployments and the specific model tables. If the subscription’s billing country/region is not in the model’s offer availability, the model cannot be deployed from that subscription.
   • For partner/community models, also check the provider’s availability and Marketplace offer rules. Errors like Offer not available in your country/region or Marketplace purchase eligibility check failed indicate that the subscription’s billing region is not eligible.
2. Ensure the Foundry project is in a supported region
   • Each model lists the allowed Hub/Project Region for Deployment. Only projects in those regions can be used. If all existing projects are in unsupported regions, the deployment dialog will show No available regions.
   • Create a new Foundry project in one of the supported regions for that model, then retry the deployment.
3. Distinguish deployment types
   • Some models are only available as Global Standard or other serverless deployment types. The Foundry project must be in a region allowed for that deployment type (for example, certain Anthropic Claude models require the project to be in East US 2 or Sweden Central).
4. Check quota and access (after regions are available)
   • Once a supported region appears and can be selected, deployment may still fail if quota is exhausted. In that case, request more quota or reallocate TPM from existing deployments.

If, after creating a project in a listed supported region, the dialog still shows No available regions for all models (including non‑GPT models), this typically indicates that none of the models in the catalog are currently offered to the subscription’s billing country/region or that Marketplace eligibility is blocking third‑party offers. In that situation, open an Azure support request so the support team can review subscription‑level availability and Marketplace eligibility for Foundry Models.

References:

• Deploy Microsoft Foundry Models in the Foundry portal
• Quota for deploying and running inference on a model
• Deploy models using Azure CLI and Bicep (programming-language-bicep)
• How to troubleshoot your deployments and monitors in Microsoft Foundry portal (classic)
• Foundry Models from partners and community
• Region availability for models in standard deployments
• Foundry Models frequently asked questions