Microsoft Foundry Ushers in New Era of Production-Ready AI Agents with Major Platform Enhancements

Microsoft is accelerating the transition of artificial intelligence from experimental concepts to tangible business value with significant updates to its Microsoft Foundry platform. Aimed at empowering organizations to build, run, govern, and distribute AI agents reliably, these enhancements mark a pivotal moment in the company’s strategy for the burgeoning agentic era. The platform now offers enhanced capabilities for integrating frontier models, robust production runtime environments, enterprise-grade security, and seamless distribution across the Microsoft 365 ecosystem. This move signifies Microsoft’s commitment to providing a unified, end-to-end solution for AI agent deployment, addressing the critical need for systems that are not only intelligent but also observable, reliable, and demonstrably aligned with business outcomes.

The announcement, made in conjunction with recent developer conferences, underscores a core promise: enabling developers to create agents within their existing workflows, leverage trusted infrastructure, and deploy them to end-users without the complexities of stitching together disparate platforms. More than 100,000 organizations are already actively building on Microsoft Foundry, with prominent companies like Adobe, Telefónica, and Tata Consultancy Services already running agents in production environments. These new, generally available updates represent a significant step in realizing this vision, transforming the platform from a roadmap into a fully functional reality.

Foundry: The Cornerstone of Production AI Agents

Microsoft Foundry is positioned as an industry-leading, end-to-end platform designed to streamline the entire lifecycle of AI agent development and deployment. Its architecture is built upon three key pillars: building, running, and governing AI agents. This integrated approach aims to eliminate the friction typically associated with moving AI solutions from proof-of-concept to large-scale production, a challenge that has historically hindered widespread AI adoption.

The platform’s emphasis on a unified experience is crucial. By consolidating the capabilities needed for frontier models, production-grade agent runtime, stringent identity and security controls, and distribution channels within Microsoft 365, Foundry empowers organizations to bypass the often-arduous process of assembling disconnected tools and services. This holistic approach is designed to foster faster innovation and more efficient deployment of AI-powered solutions across diverse business functions.

Empowering Developers with Flexible and Powerful Tools

A central tenet of Foundry’s strategy is meeting developers where they work. Agent development is integrated directly into familiar environments such as GitHub Copilot and Microsoft Visual Studio (VS) Code. The Foundry Toolkit for VS Code and the Foundry skill facilitate seamless deployment to the Foundry platform. This approach supports a variety of development frameworks, including Microsoft Agent Framework and the GitHub Copilot SDK, as well as emerging options like the Claude Agent SDK. Regardless of the chosen framework, Foundry serves as the robust production destination.

The power of an AI agent is intrinsically linked to the underlying models that drive its reasoning capabilities. Microsoft Foundry addresses this by providing access to a diverse range of industry-leading models, including frontier, open-source, and task-specific options, all through a single, cohesive platform. This allows development teams to select the most appropriate model for each specific workload, optimizing for capability, cost, and performance.

Introducing the GPT-5.6 Series: Enhanced Performance and Accessibility

A significant highlight of the recent updates is the general availability of OpenAI’s GPT-5.6 series within Microsoft Foundry Models and Microsoft Foundry Agent Service. This new series offers organizations enhanced flexibility, enabling them to tailor model selection to specific business scenarios rather than being constrained by a single, generalized model.

The GPT-5.6 series is structured into three tiers: Sol, Terra, and Luna, each offering distinct price points and performance characteristics. GPT-5.6 Sol is positioned for maximum capability, while GPT-5.6 Terra provides a balance of performance and cost-effectiveness. GPT-5.6 Luna is designed for cost-optimized workloads where efficiency is paramount. This tiered approach allows businesses to align their AI investments precisely with their needs, fostering more efficient resource allocation and maximizing return on investment.

Microsoft is ensuring that these advanced models are accessible across its global infrastructure. The GPT-5.6 series is available through Global Standard and Global Priority Processing in all 28 global regions, as well as Data Zones Standard and Global Provisioned from day one. This broad availability ensures that customers can adopt the latest AI innovations within their existing deployed applications and infrastructure, regardless of their geographic location.

The pricing structure for the GPT-5.6 series is as follows, denominated in USD per million tokens:

This transparent pricing model empowers organizations to conduct thorough cost-benefit analyses and plan their AI deployments with greater financial certainty.

Expanding Global Reach with the APAC Data Zone

Beyond model availability, Microsoft Foundry is enhancing its operational footprint to cater to diverse regulatory and data sovereignty requirements. The general availability of the Asia-Pacific (APAC) Data Zone for Microsoft Foundry is a significant development. This new zone enables customers in the Asia-Pacific region to run frontier OpenAI models while ensuring that data processing remains within regional boundaries. This eliminates the need for complex, separate environments and accelerates the adoption of advanced AI capabilities without compromising compliance.

With the introduction of the APAC Data Zone, Foundry now offers Global, Data Zone, and Regional deployment options. This flexibility allows organizations to align their AI strategies with specific sovereignty, compliance, performance, and scalability mandates, all while maintaining a consistent and familiar development and operations experience across different environments.

Hongsoo Kim, Chief Data and AI Officer (CDAO) at Viva Republica (Toss), highlighted the importance of such regional capabilities for financial institutions: "As financial institutions adopt AI, responsible data handling becomes foundational to trust. Microsoft Foundry’s APAC Data Zone allows us to keep data processing regionally anchored while accessing advanced AI models at scale. This gives us the confidence to accelerate AI innovation responsibly and reinforces our ambition to be a leading AI-powered financial platform in Asia."

Building Action-Oriented, Context-Aware Agents

A powerful model is merely the foundation. To be effective in production, an AI agent requires a comprehensive ecosystem of capabilities. Microsoft Foundry provides these built-in functionalities as integrated components, designed to work harmoniously. These include a robust runtime environment for agents to execute, deep integration of business knowledge, secure and governed access to essential tools, persistent memory across interactions, and the ability to act upon real-world events. Furthermore, Foundry ensures a clear pathway for agents to reach the end-users who will leverage their capabilities.

The platform’s comprehensive suite of features for agent development includes:

• Agent Orchestration: Enables complex agent workflows and interactions.
• Tool Integration: Seamlessly connects agents with external applications and services.
• Knowledge Retrieval: Allows agents to access and utilize relevant organizational data.
• Memory Management: Provides agents with context and continuity across conversations.
• Event Handling: Enables agents to respond to real-time triggers and events.
• Distribution: Facilitates deployment and access across Microsoft 365 applications.

Governing and Optimizing the AI Lifecycle

Trust is paramount in AI deployments, and Microsoft Foundry treats it as a core platform priority, not solely a developer responsibility. The latest enhancements focus on providing comprehensive observability and control over the entire AI lifecycle, extending beyond the initial build phase. This includes detailed insights into agent actions, mechanisms for continuous improvement, and verifiable proof of an agent’s value and operational integrity.

Key features for governing and optimizing AI agents include:

• Observability: Comprehensive logging and monitoring of agent behavior and performance.
• Evaluation: Tools to assess agent effectiveness and identify areas for enhancement.
• Tracing: Detailed tracking of agent requests and responses for debugging and auditing.
• Security and Compliance: Robust controls to ensure data privacy and adherence to regulatory standards.
• Cost Management: Tools to monitor and optimize AI spending effectively.

As agents scale from pilot projects to handling thousands of daily requests, Foundry equips teams with the tools to manage costs predictably without leaving the platform. This cost optimization is achieved through a combination of strategic features:

• Model Router: Intelligently matches each request to the most appropriate and cost-effective model.
• Prompt Caching: Reduces redundant computations by storing and reusing common prompts.
• PTU Spillover and Quota Optimization: Ensures service continuity during usage spikes and efficient quota management.
• Toolbox Optimization: Ensures agents only utilize the specific tools required for a given request, minimizing overhead.
• Agent Optimizer: Tunes prompts, skills, tools, and model choices against custom evaluators to maximize efficiency and performance.

Furthermore, Foundry provides a clear view of an agent’s return on investment (ROI) by correlating business value, usage metrics, and operational costs. This integrated financial perspective enables teams to ascertain whether their production agents are generating more value than they incur in costs, and to identify areas where cost optimization may be needed. A detailed examination of token economics for agents is available in a recent Microsoft Mechanics episode.

Real-World Impact: Organizations Building on Foundry

The adoption of Microsoft Foundry is not confined to experimentation; it is actively enabling organizations across the spectrum, from digital natives to global enterprises, to deploy production-ready AI solutions. Companies are reporting a significant acceleration in their AI deployment timelines, with integration, security, and deployment processes that once took weeks now being completed in days. This efficiency, combined with the assurance of operating on compliant infrastructure and reaching users through familiar tools, is driving tangible business outcomes.

Getting Started with Microsoft Foundry

All the functionalities detailed in this announcement are now live and accessible within Microsoft Foundry. Comprehensive documentation and Microsoft Learn courses are available to guide users through the platform’s capabilities. Developers can quickly begin their journey by following the Quickstart guide, which provides an end-to-end walkthrough of setting up, testing, and deploying a production-ready hosted agent.

For those seeking structured learning, the "AI Agents for Beginners" curriculum offers a 12-lesson series. Deeper dives are available through guided labs such as "Develop AI Agents in Azure," "Hosted Agents Workshop (.NET)," the "Foundry Toolkit for VS Code and hosted agents workshop," and the "ZavaShop Supply Chain Workshop." To ensure the quality and reliability of AI agents, "Evaluating AI Agents: A Practical Guide with Microsoft Foundry" offers valuable insights.

A comprehensive overview of operationalizing AI agents, from deployment to real-world impact, is provided in the Microsoft Mechanics episode "Foundry Agent Service + Microsoft Agent Framework Explained," featuring Jeff Hollan. These resources collectively aim to empower developers and organizations to harness the full potential of AI agents within a secure, scalable, and production-ready environment.