AI Foundry Models
Azure AI Foundry Models let you access AI models directly through your Azure subscription, from OpenAI and other providers like Cohere, DeepSeek, Meta, Mistral AI, and xAI. You deploy these models to a Foundry resource and use them through REST APIs or SDKs, with everything running under Azure's standard service-level agreements and unified billing.
Details
Azure AI Foundry Models are a collection of AI models you can use directly through your Azure subscription. This includes all the Azure OpenAI models, plus models from other top providers like Cohere, DeepSeek, Meta, Mistral AI, and xAI. Everything runs under Azure's service-level agreements and support.
When you're ready to use a model, you deploy it to a Foundry resource. You have two main deployment types to choose from. Standard deployments give you global deployment with higher throughput, which works well for most everyday applications. Provisioned deployments let you purchase dedicated throughput units when you need guaranteed performance for production workloads or high-traffic scenarios.
