Microsoft Foundry Unveils GPT-5.6, Asia-Pacific Data Zone, and Hosted Agents, Ushering in New Era of Production-Ready AI

Microsoft Foundry is now generally available with a suite of powerful new features, including the highly anticipated GPT-5.6 series, the new Asia-Pacific Data Zone, and the integration of hosted agents within the Foundry Agent Service. This significant expansion marks a pivotal moment in the democratization of advanced AI, empowering over 100,000 organizations already building on the platform to move beyond experimentation and into reliable, observable, and business-outcome-aligned production deployments. Industry leaders like Adobe, Telefónica, and Tata Consultancy Services are already leveraging these agentic capabilities in their live operations, demonstrating the tangible value and scalability of Microsoft’s AI ecosystem.

The announcements, made with a clear promise to developers to build agents where they work, run them on trusted infrastructure, and distribute them to end-users without platform fragmentation, transition from roadmap to reality today. These updates collectively integrate frontier models, robust production agent runtimes, enterprise-grade identity and security controls, and seamless distribution across Microsoft 365 into a unified platform. This consolidation aims to eliminate the need for complex stitching of disparate tools and services, streamlining the journey from initial concept to full-scale production.

Microsoft Foundry is positioned as the industry’s premier end-to-end platform for the entire AI agent lifecycle: building, running, governing, and distributing. The platform’s strength lies in its three core pillars:

• Foundry Models: Providing access to a diverse range of cutting-edge, open-source, and specialized models.
• Foundry Agent Service: Offering a secure and scalable runtime environment for deploying and managing AI agents.
• Foundry Distribution: Enabling broad reach and accessibility of agents across Microsoft’s expansive ecosystem.

These pillars are brought to life by the latest Foundry updates, which are designed to facilitate the building, running, and scaling of production-ready agents on a singular, cohesive platform.

Building with Flexibility: Any Framework, Any Model

Agent development within Microsoft Foundry begins at the familiar developer workstations, namely GitHub Copilot and Microsoft Visual Studio (VS) Code. The Foundry Toolkit for VS Code and the Foundry skill streamline the deployment process to Foundry. Whether teams opt for the Microsoft Agent Framework, the GitHub Copilot SDK (now generally available), or the Claude Agent SDK, Foundry serves as the ultimate production destination. The foundation of any capable agent lies in its underlying model, and Microsoft Foundry excels in providing access to a broad spectrum of industry-leading models.

The Right Model for Every Task: Introducing GPT-5.6 Series

The efficacy of an AI agent is directly proportional to the reasoning power of its underlying model. Microsoft Foundry democratizes access to a curated selection of frontier, open-source, and task-specific models, enabling organizations to meticulously match model capabilities, costs, and performance metrics to their unique workloads.

Today’s general availability of OpenAI’s GPT-5.6 series within Microsoft Foundry Models and Microsoft Foundry Agent Service represents a significant leap forward. This series includes:

• GPT-5.6 Sol: The most powerful and versatile model in the series, designed for complex reasoning and nuanced tasks.
• GPT-5.6 Terra: A balanced model offering strong performance across a wide range of applications, optimized for efficiency.
• GPT-5.6 Luna: A cost-effective and highly performant model, ideal for high-volume, less complex tasks.

This tiered approach grants organizations the strategic flexibility to align model selection with specific business scenarios, moving away from a one-size-fits-all approach. This granular control over model choice is crucial for optimizing both performance and cost-effectiveness.

Customers have consistently emphasized that access to the latest models is as vital as model quality itself. To address this, Microsoft is making the GPT-5.6 series available across 28 global regions, including Standard and Priority Processing options, as well as Data Zones Standard and Global Provisioned deployments from day one. This ensures that organizations can immediately leverage cutting-edge AI innovations within their existing, established application infrastructures.

GPT-5.6 Pricing Structure (USD per million tokens):

Note: Pricing for Data Zones and Global Provisioned deployments may vary.

Global Reach, Local Compliance: The Asia-Pacific Data Zone

Expanding access to advanced AI models is only one facet of platform evolution; ensuring compliant execution in diverse geographical locations is equally critical. The newly available Asia-Pacific Data Zone for Microsoft Foundry addresses this need directly. This expansion allows APAC customers to utilize frontier OpenAI models while ensuring that data processing remains within the Asia-Pacific regions. This eliminates the need for complex, separate environment configurations and the wait for feature parity.

With comprehensive deployment options—Global, Data Zone, and Regional—Foundry empowers organizations to align their AI adoption strategies with their specific sovereignty, compliance, performance, and scalability requirements. This is all delivered through a consistent and unified development and operational experience across all environments.

Hongsoo Kim, Chief Data and AI Officer at Viva Republica (Toss), commented on the significance of this development: "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." This sentiment underscores the critical importance of regional data control for sensitive industries.

Action-Oriented Agents: Beyond Model Capabilities

A powerful model is merely the initial building block. To translate that power into production-ready applications, an AI agent requires a robust operational framework. This includes a reliable runtime environment, deep understanding of business context, governed access to necessary tools, persistent memory across interactions, the ability to act on real-world events, and a clear pathway to end-users. Foundry integrates all these essential components as built-in capabilities, meticulously designed to function harmoniously.

The core components enabling agents to generate tangible business impact include:

• Hosted Agents: Providing a secure, scalable, and managed environment for agent execution.
• Agent Memory: Enabling agents to retain context and learn from past interactions, fostering more intelligent and personalized experiences.
• Agent Tooling: Granting agents secure and governed access to a wide array of internal and external tools and APIs.
• Agent Orchestration: Managing the complex workflows and interactions between different agent components and external systems.
• Agent Distribution: Facilitating the seamless deployment and accessibility of agents across various user interfaces and applications, including Microsoft 365.

Governing the AI Lifecycle: Observability and Control

An AI agent that cannot be monitored, improved, or secured is an agent that cannot be safely deployed into production. Microsoft Foundry prioritizes trust as a foundational platform tenet, rather than placing the burden solely on developers. The latest release enhances the post-build lifecycle, offering unparalleled visibility into agent actions, facilitating continuous improvement, and providing verifiable proof of their value.

Key enhancements for governing and optimizing the AI lifecycle include:

• Agent Observability: Comprehensive logging and monitoring provide deep insights into agent behavior, performance, and resource utilization. This allows for proactive identification of issues and optimization opportunities.
• Agent Evaluation: Tools and frameworks for rigorously testing and validating agent performance against predefined metrics and business objectives, ensuring reliability and accuracy.
• Agent Security: Robust security controls, including identity management, access policies, and data encryption, safeguard agent operations and sensitive data.
• Cost Management: Granular cost tracking and optimization tools allow organizations to monitor and control AI expenditures as agents scale from pilot programs to thousands of daily operations.

Foundry empowers teams to maintain predictable spending without leaving the platform, starting with intelligent choice. This includes:

• Model Routing: The integrated Model Router intelligently matches each request to the most appropriate model, balancing performance, cost, and capability.
• Prompt Caching: This feature significantly reduces redundant computation by storing and reusing results from identical prompts, thereby improving efficiency and lowering costs.
• PTU Spillover and Quota Optimization: These mechanisms ensure service continuity during peak usage spikes, preventing performance degradation and maintaining accessibility.
• Toolbox Optimization: Agents are equipped with dynamic toolboxes that provide only the necessary tools for each specific request, reducing overhead and improving response times.
• Agent Optimizer: This advanced tool fine-tunes prompts, skills, tools, and model choices by evaluating them against the organization’s custom evaluators, leading to more effective and efficient agents.

Beyond cost optimization, Foundry offers a clear view of Return on Investment (ROI) for agents. This feature integrates business value, usage metrics, and operational costs into a single, consolidated dashboard. This allows teams to definitively assess whether a production agent is generating more value than it incurs in costs, and to pinpoint areas where cost may be outpacing value.

For a practical understanding of these token economics, the "Microsoft Mechanics" episode on the subject provides an insightful walkthrough.

Real-World Impact: Production Deployments on Foundry

The organizations actively building on Microsoft Foundry are not merely experimenting; they are successfully deploying AI agents into production environments. This spans a wide spectrum, from agile digital natives to established global enterprises.

Companies like Adobe, a leader in creative and digital experiences, are leveraging Foundry to enhance their customer engagement and operational efficiency. Telefónica, a telecommunications giant, is integrating AI agents to optimize customer service and network management. Tata Consultancy Services (TCS), a premier IT services and consulting company, is utilizing Foundry to build and deploy AI solutions for its diverse global clientele, accelerating digital transformation initiatives.

The consistent pattern observed is that teams who previously spent weeks on the complex integration, security, and deployment of AI agents can now achieve similar results in days. This accelerated deployment is enabled by an infrastructure that meets stringent compliance requirements and delivers agents through user-trusted applications.

Getting Started with Microsoft Foundry

All the features and capabilities discussed are live and accessible within Microsoft Foundry. Aspiring developers and organizations can begin their journey by consulting the comprehensive documentation and Microsoft Learn courses. A streamlined Quickstart guide is available for developers to set up, test, and deploy a production-ready hosted agent from end to end in minutes.

Further educational resources include:

• AI Agents for Beginners: A 12-lesson curriculum designed to introduce foundational concepts.
• Guided Labs: Including "Develop AI Agents in Azure," "Hosted Agents Workshop (.NET)," "Foundry Toolkit for VS Code and hosted agents workshop," and the "ZavaShop Supply Chain Workshop."
• Practical Guide: "Evaluating AI Agents: A Practical Guide with Microsoft Foundry" offers best practices for ensuring agent quality.

For a detailed explanation of how to operationalize AI agents from deployment to real-world impact, watch Jeff Hollan’s walkthrough in the "Foundry Agent Service + Microsoft Agent Framework Explained" video. This comprehensive suite of resources aims to lower the barrier to entry and accelerate the adoption of production-ready AI agents across industries.