Skip to main content
Microsoft Foundry provides customers with choices on the hosting structure that fits their business and usage patterns. In Foundry Models sold by Azure, model offerings include both Azure OpenAI models and models from other providers. This article introduces regional availability across both collections. Serverless API deployment in Microsoft Foundry includes three deployment categories: standard (pay-per-token), provisioned (reserved capacity), and batch (for asynchronous requests). Within these categories, you can choose global, data zone, or regional deployment types based on your compliance requirements. For all deployment types, data stored at rest remains in the designated Azure geography (Americas, Europe, Asia Pacific, and Middle East and Africa). However, inferencing data is processed as follows:
  • Global types: Might be processed in any Azure region where the model is deployed
  • Data Zone types: Processed anywhere within the Microsoft-specified data zone (US, EU, or Asia Pacific (APAC))
  • Standard/Regional types: Processed in the region associated with your deployment (not available for batch deployments)
All deployments can perform the exact same inference operations, but the billing, scale, and performance are substantially different. To learn more about Microsoft Foundry deployment types, including batch deployment types, see Deployment types for Microsoft Foundry Models.
Use the tabs at the top of this page to switch deployment categories: Standard deployment options, Provisioned deployment options, and Batch deployment options.

​
Related content