Formerly known as Azure AI Services or Azure Cognitive Services is a unified collection of prebuilt AI capabilities within the Microsoft Foundry platform
Hi Karan Agrawal,
This issue usually happens due to subscription, billing, or region limitations, not because of a problem with your code or project setup.
Below are a few things you can check to understand why the Claude model deployment is failing.
- Please check your Azure subscription type. Anthropic Claude models in Azure AI Foundry require a paid Azure subscription with an active billing method. Student, free trial, sponsored, or credit‑only subscriptions cannot deploy Claude models. This is expected behavior and not a quota bug.
You can find this documented here https://learn.microsoft.com/azure/foundry/foundry-models/how-to/use-foundry-models-claude
- Verify that Azure Marketplace is enabled for your subscription. Claude models are offered through Azure Marketplace. If Marketplace purchasing is blocked at the subscription or tenant level, deployment will fail even if everything else looks correct.
- Confirm that your billing profile details are complete. If your billing account is missing organization name, address, or country information, the Marketplace validation can fail during deployment. This is a common reason when the portal shows deployment failure without a clear error.
- Make sure you are deploying in a supported region. Claude models are currently supported only in specific regions like East US 2 and Sweden Central. If your Foundry project is created in another region, the deployment will not work.
Supported regions are listed in the official documentation https://learn.microsoft.com/azure/foundry/foundry-models/how-to/use-foundry-models-claude
Also note that Anthropic models do not appear in the standard Azure AI quota page. This is expected because they are Marketplace based models. Seeing zero or missing quota for Claude does not always mean access is blocked.
Once the above conditions are met, deployment usually works without any changes to your code.
Hope this helps clarify the behavior and saves you some time.
Thankyou!