Azure API Management: Setup MCP Servers

Key insights

• Azure API Management now lets you expose REST APIs as MCP servers, making it easy for AI agents and large language models to connect with your existing APIs using the new Model Context Protocol (MCP).


• The MCP protocol is an open standard based on JSON-RPC, designed so that AI-powered agents can call external tools and APIs in a consistent and secure way.


• This technology allows for centralized control of authentication, authorization, and monitoring, which helps keep your APIs secure while scaling their use in AI-driven workflows.


• API Center acts as a single place to discover and manage all MCP servers within an organization, improving enterprise API governance and visibility.


• You can expose current REST APIs as MCP servers without rewriting code, so developers can quickly connect these APIs to AI workflows and test them through MCP clients.


• This approach bridges the gap between stateless AI models and real-world data by standardizing how AI agents interact with external services, while also adding enterprise-level security and management features.

Introducing Azure API Management’s New MCP Integration

Microsoft Azure Developers have introduced a significant update to Azure API Management, enabling users to expose REST APIs as Model Context Protocol (MCP) servers. This innovation, showcased in a recent YouTube video session led by Julia Kasper, marks a turning point for organizations aiming to connect their APIs with advanced AI agents such as GitHub Copilot. The technology, announced in May 2025 and currently available in preview, promises a streamlined approach to integrating REST APIs into AI-driven workflows, all without requiring developers to write new code.

At the heart of this update is MCP, a JSON-RPC-based open standard that lets AI-powered agents and large language models (LLMs) invoke external tools and APIs in a standardized manner. By leveraging Azure API Management, enterprises can now make their APIs accessible as MCP servers, thus providing AI systems with real-world data and services on demand.

Key Benefits and Developer Advantages

One of the most compelling aspects of this new feature is its ability to deliver secure and scalable AI integration. Azure API Management grants centralized control over authentication, authorization, and monitoring for all MCP servers. This centralized approach not only reduces security risks but also ensures that organizations can scale their API usage efficiently as AI workloads grow.

Additionally, MCP servers are managed much like traditional backend APIs within Azure API Management, allowing teams to apply familiar governance, security, and operational controls. The introduction of the API Center further enhances this experience by acting as a unified registry where enterprises can discover, catalog, and manage all their MCP servers. Developers benefit from the seamless process of exposing existing REST APIs as MCP servers, eliminating the need for extensive rewriting or complex integration work.

The Technology Explained: How It Works

The foundation of this approach lies in the Model Context Protocol (MCP), which standardizes the communication between AI agents and external APIs using JSON-RPC. Once a REST API is managed by Azure API Management, specific operations can be exposed as MCP tools. Organizations can then configure policies within API Management to define how these MCP servers behave, including vital security and operational settings.

Testing is also straightforward: once an MCP server is set up, it can be verified using specialized MCP clients to ensure proper connectivity and functionality. Since the feature is still in preview, it is available initially through the AI Gateway Early update group, with new users typically experiencing a short delay before activation. The preview phase allows Microsoft to gather feedback and make improvements before a broader release.

Novelty and Tradeoffs of the MCP Approach

What sets this solution apart is its ability to bridge the gap between stateless AI models and external APIs using a standardized protocol. Traditionally, integrating AI agents with APIs required custom or ad-hoc solutions, which could be difficult to manage and scale. By adopting MCP, organizations can now expose their APIs as standardized tools, making them easily consumable by AI agents while maintaining robust security and management practices.

However, there are tradeoffs to consider. While the approach simplifies integration and governance, organizations must ensure their existing APIs are compatible with the MCP framework. Some may need to adjust policy configurations or invest time in learning new management practices. Additionally, as the feature is still in preview, early adopters might encounter evolving documentation and occasional limitations as Microsoft continues to refine the service.

Resources and Future Outlook

To support developers and organizations, Microsoft provides comprehensive documentation detailing how to expose REST APIs as MCP servers, configure necessary policies, and test integrations. The recent Microsoft Build 2025 conference also featured live demonstrations, helping users visualize the end-to-end process from setup to deployment.

Looking ahead, as MCP matures and becomes generally available, it is expected to play a key role in enabling secure, scalable, and standardized AI integrations across enterprises. By leveraging existing infrastructure and familiar API management tools, organizations can confidently extend their AI capabilities while maintaining high standards for security and governance.

Keywords

Supercharge APIs Azure API Management REST APIs MCP Servers API exposure API gateway cloud integration developer tools