Azure AI Foundry Model Router

Microsoft Foundry offers a comprehensive platform-as-a-service for enterprise AI operations. It provides multiple deployment options including Serverless APIs (pay-as-you-go), Global Standard (shared managed capacity), Provisioned Throughput Units (reserved capacity), batch processing, and bring-your-own model deployments. The platform features a unified control plane for models, agents, tools, and observability. Its Agent Service enables building and deploying AI agents with built-in tracing, monitoring, and governance. Evaluation and monitoring tools assess model performance, safety, and groundedness. Foundry supports seamless upgrades from Azure OpenAI with non-destructive migration, maintaining existing deployments while unlocking multi-provider model access and advanced platform capabilities.

Azure OpenAI Service hosts OpenAI's GPT-4o, GPT-4, GPT-3.5, and embedding models on Microsoft Azure with enterprise SLAs. Deployments run in customer-selected regions with private networking, role-based access control, and capacity options spanning Standard pay-per-token, Provisioned Throughput Units (PTUs) for reserved capacity, Global Standard shared capacity, and Batch processing. Azure OpenAI sits inside the wider Microsoft Foundry / Azure AI Studio control plane, which adds an evaluation, monitoring, and Agent Service layer on top of the base model APIs. For workloads that need non-OpenAI models (Claude, DeepSeek, Grok, Llama, Mistral, NVIDIA Nemotron), Microsoft Foundry is the broader catalog; Azure OpenAI is the OpenAI-specific entry point. The service is API-compatible with the OpenAI SDK in most flows, so teams typically swap base URLs and authentication rather than rewriting calls.