Introducing GPT-5.4 in Microsoft Foundry

Built for Reliable AI Production: Stronger reasoning, dependable execution, and agentic workflows at scaleToday, we’re announcing OpenAI’s GPT‑5.4 to be generally available soon in Microsoft Foundry: a model designed to help organizations move from planning work to reliably completing it in production environments. As AI agents are applied to longer, more complex workflows; consistency and follow‑through become as important as raw intelligence. GPT‑5.4 combines stronger reasoning with built in computer use capabilities to support automation scenarios, and dependable execution across tools, files, and multi‑step workflows at scale.

GPT-5.4: Enhanced Reliability in Production AIGPT-5.4 is designed for organizations operating AI in real production environments, where consistency, instruction adherence, and sustained context are critical to success. The model brings together advances in reasoning, coding, and agentic workflows to help AI systems not only plan tasks but complete them with fewer interruptions and reduced manual oversight.

Compared with earlier generations, GPT-5.4 emphasizes stability across longer interactions, enabling teams to deploy agentic AI with greater confidence in day-to-day production use.

GPT-5.4 introduces advancements that aim for production grade AI:

More consistent reasoning over time, helping maintain intent across multi‑turn and multi‑step interactionsEnhanced instruction alignment to reduce prompt tuning and oversightLatency improved performance for responsive, real-time workflowsIntegrated computer use capabilities for structured orchestration of tools, file access, data extraction, guarded code execution, and agent handoffsMore dependable tool invocation reducing prompt tuning and human oversightHigher‑quality generated artifacts, including documents, spreadsheets, and presentations with more consistent structureTogether, these improvements support AI systems that behave more predictably as tasks grow in length and complexity.

From capability to real-world outcomesGPT‑5.4 delivers practical value across a wide range of production scenarios where follow‑through and reliability are essential:

Agent‑driven workflows, such as customer support, research assistance, and business process automationEnterprise knowledge work, including document drafting, data analysis, and presentation‑ready outputsDeveloper workflows, spanning code generation, refactoring, debugging support, and UI scaffoldingExtended reasoning tasks, where logical consistency must be preserved across longer interactionsTeams benefit from reduced task drift, fewer mid‑workflow failures, and more predictable outcomes when deploying GPT‑5.4 in production.

GPT-5.4 Pro: Deeper analysis for complex decision workflowsGPT‑5.4 Pro, a premium variant designed for scenarios where analytical depth and completeness are prioritized over latency.

Additional capabilities include:

Multi‑path reasoning evaluation, allowing alternative approaches to be explored before selecting a final responseGreater analytical depth, supporting problems with trade‑offs or multiple valid solutionsImproved stability across long reasoning chains, especially in sustained analytical tasksEnhanced decision support, where rigor and thoroughness outweigh speed considerationsOrganizations typically select GPT‑5.4 Pro when deeper analysis is required such as scientific research and complex problems, while GPT‑5.4 remains the right choice for workloads that prioritize reliable execution and agentic follow‑through.

Microsoft Foundry: Enterprise‑Grade Control from Day OneGPT‑5.4 and GPT‑5.4 Pro are available through Microsoft Foundry, which provides the operational controls organizations need to deploy AI responsibly in production environments. Foundry supports policy enforcement, monitoring, version management, and auditability, helping teams manage AI systems throughout their lifecycle.

By deploying GPT‑5.4 through Microsoft Foundry, organizations can integrate advanced agentic capabilities into existing environments while aligning with security, compliance, and operational requirements from day one.

Customer Spotlight

Get Started with GPT-5.4 in Microsoft FoundryGPT‑5.4 sets a new bar for production‑ready AI by combining stronger reasoning with dependable execution. Through enterprise‑grade deployment in Microsoft Foundry, organizations can move beyond experimentation and confidently build AI systems that complete complex work at scale. Computer use capabilities will be introduced shortly after launch.

GPT‑5.4 in Microsoft Foundry is priced at $2.50 per million input tokens, $0.25 per million cached input tokens, and $15.00 per million output tokens. It is available at launch in Standard Global and Standard Data Zone (US), with additional deployment options coming soon. GPT‑5.4 Pro is priced at $30.00 per million input tokens, and $180.00 per million output tokens, and is available at launch in Standard Global.

Build agents for real-world workloads. Start building with GPT‑5.4 in Microsoft Foundry today.
The post Introducing GPT-5.4 in Microsoft Foundry appeared first on Microsoft Azure Blog.
Quelle: Azure

The economics of enterprise AI: What the Forrester TEI study reveals about Microsoft Foundry

Leaders are chasing the AI frontier, reimagining business systems as human-led and agent-operated. To do this, customers are on the hunt for smarter models, more capable agents, and market-ready solutions to operationalize AI workflows.

When Forrester modeled the economics of enterprise AI with Microsoft Foundry, the biggest driver behind the 327% ROI over three years1 was surprising: developer productivity, worth $15.7 million over the same period.

The study showed that the bottleneck to ROI can be removed by enabling developers to focus on what matters.

Read the full Forrester study

The hidden tax on your AI investment

In most organizations, senior engineers spend a third of their time on undifferentiated work: stitching together fragmented tools, recreating context pipelines, and navigating bespoke governance processes. None of that is competitive advantage for firms—it’s a tax on every AI initiative.

According to Forrester, organizations using Foundry avoided much of this work, improving technical team productivity up to 35%. Teams using Foundry to develop AI apps and agents saw payback in as few as six months and with benefits accelerating year over year1.

Learn more about what you can do with Microsoft Foundry

The details: What the Forrester study found

Forrester interviewed 10 decision-makers at five organizations and surveyed 154 other decision-makers and AI leaders across the U.S. and Europe with experience using Microsoft Foundry. They modeled a composite enterprise with $10 billion revenue, 25,000 employees, and 100 technical staff using Foundry. To model conservative estimates, benefits were adjusted downward and costs upward; the results reflect the composite enterprise.

Read the full Forrester study

Figure 1: Survey results and reported benefits

When asked “What benefits has your organization experienced with Microsoft Foundry?”, respondents cited operational outcomes:

Note: These reflect reported experiences, not the financial model. Composite ROI is calculated separately using Forrester’s risk-adjusted methodology. Source: Survey of 154 AI decision-makers, Forrester TEI study, February 2026

Forrester found that platform investments compound in value. For a team that invests $11.6M in resources, the three-year present value of quantified benefits for the composite organization totaled $49.5M: Year one delivered $10.0M, year two $21.1M, year three $30.5M.

Figure 2: Benefits breakdown

Source: The Total Economic Impact™ Of Microsoft Foundry, a commissioned study conducted by Forrester Consulting, February 2026

When every project starts from scratch

AI initiatives will require models, enterprise knowledge, tools, and governance. Without a shared platform, teams will encounter toil. With enterprise knowledge as the example, for every AI project, teams need to create vector databases, RAG pipelines, integrations, and access-control rules, creating internal infrastructure that does not directly influence business outcomes.

75% of teams reported easier model grounding or knowledge source integration

Read the study

With Foundry, teams develop AI applications and agents on a unified, interoperable AI platform designed to enable agents to be intelligent and trustworthy: with reusable knowledge bases on data anywhere in the enterprise, protected by built-in evaluations, and agent controls. In Forrester’s TEI study, 75% of teams cited easier model grounding or knowledge source integration with Foundry IQ.

Over three years, the productivity gain alone was worth up to $15.7 million1. One Foundry customer said,

Our developers can go super fast because they can get what they need in Microsoft Foundry … We estimate that we reduce overall development time by 30%–40%.
—Global head of technology platforms, professional services

Organizations saw compounding returns when they built once and reuse everywhere with shared templates, knowledge bases, standardized evaluations, and consistent governance. This helps to explain a counterintuitive finding: organizations that focused energy consolidating on a unified platform outperformed those which did not. Their execution is simpler and therefore stronger.

The need for platform thinking

Point solutions develop in enterprises over time. Each solves a narrow problem, but each also introduces its own governance layer, context pipeline, and integration surface. The hidden cost here builds up in the stitching between these solutions.

32% were able to decrease costs by decommissioning legacy AI tools

Read the study

In the Forrester study, 32% of surveyed organizations that adopted Foundry were able to decrease costs by decommissioning legacy AI tools, and the composite organization avoided up to $4.3M in infrastructure costs over three years by eliminating duplicative workflows, integrations, and operational overhead. For example, one customer shared they were able to decommission their container-based infrastructure and eliminate spending on previous AI model development tools since the functionality was included in the Foundry platform:

One of the benefits of using Foundry versus taking those models and running them in containers in the cloud is that then you don’t have to manage the container infrastructure.
—Managing director and global head of co-innovation, professional services

Department-level budgets favor point solutions, but enterprise-level outcomes require platform thinking. That mismatch is why AI spend often fails to translate into sustained value as organizations shift from isolated pilots to scaled deployments.

Microsoft Agent Factory
Scale AI and move from ideas to outcomes with one pre-paid plan, expert-led AI skilling, and engineering expertise.

Learn more

Trust unlocks higher-impact work

Most enterprises start with internal-facing AI use cases before they shift to customer-facing solutions. Two-thirds of AI agents today focus on process automation, while one-third support direct human assistance1. The ratio matters. Most enterprises need to trust AI with bounded, auditable tasks before they can trust it to enhance human judgment.

Foundry Control Plane enables organizations to govern the AI lifecycle with organization-wide observability and controls. This includes centrally managed policies for model deployment, configurable guardrails, and continuous evaluations to see what’s running, fix what’s failing, and prove compliance across any environment.

Model scanning done by Microsoft on the models … is a key requirement for us. … we want to make sure we understand what the model contains and whether it contains anything that is not in line with policy.
—Principal product manager, professional services

67% adopted Foundry to reduce concerns with AI security, privacy, and governance

Read the study

It’s no surprise that 67% of surveyed organizations cited concerns with AI security, privacy, or governance as a top reason for adopting Microsoft Foundry, ranking it higher than model access, capabilities, and cost inefficiencies. In essence, trust is a permission slip that enables organizations to expand from isolated process automation projects into higher-impact work at scale.

What leaders should do about AI now

The Forrester TEI study makes one thing unmistakable: enterprise AI ROI compounds when AI is treated as a platform, not a series of one-off projects.

The biggest gains come from giving technical teams a reusable foundation, including models, agents, and tools that scale across use cases and eliminate repetitive work. When AI development becomes repeatable, value accelerates and confidence follows.

Three questions for your next leadership meeting
– How much of your engineering capacity goes toward rebuilding the same foundations vs. building differentiated AI capabilities? If it’s over 20%, you’re paying a hidden tax.– Do your AI initiatives share a common platform for data, evaluation, and governance, or are you scaling fragmentation?– What would it take for your organization to move from isolated automation projects to higher‑impact use cases?

Learn more about the benefits of AI workflows

Read the full Forrester TEI Study.

Build with Microsoft Foundry.

Shift from ideas to outcomes faster with Microsoft Agent Factory.

Read the full Forrester study

The Forrester Total Economic Impact™ study on Microsoft Foundry was commissioned by Microsoft and conducted by Forrester Consulting.

1The Total Economic Impact™ Of Microsoft Foundry, a commissioned study conducted by Forrester Consulting, February 2026

2Represents results for the composite organization
The post The economics of enterprise AI: What the Forrester TEI study reveals about Microsoft Foundry appeared first on Microsoft Azure Blog.
Quelle: Azure

Unpacking your top questions on agentic AI: The Shift podcast

Every day in the hallways at Microsoft, I hear product teams discussing where agents are headed and how software is forever changed. Many of us come into the office more now, and I didn’t realize how much I missed the in-between moments where natural chat gives us energy—coffee and hot takes on the way to meetings and debating at a lunch no one scheduled, but somehow nobody wants to leave. The people who work on Microsoft Azure, Microsoft Foundry, and Microsoft Fabric care deeply about what they’re building—about how cloud and AI platforms can be better for those with hands on keyboards—it’s when we’re unscripted that some of our best insights surface. How could we bottle up this passion?

Subscribe to The Shift podcast

Today we’re introducing “The Shift” podcast, an evolution of “Leading the Shift,” to share more dialogue. Grounded in questions we heard from you after announcements at Ignite, we’re releasing eight episodes this spring—one each week—that bring engineering, product, and strategy perspectives together. Across levels and backgrounds, this season’s agentic theme explores agents up and down the stack. Knowing change is the only constant, “The Shift” creates space for us all to think out loud.

Here’s a sneak peek of the new season

const currentTheme =
localStorage.getItem(‘msxcmCurrentTheme’) ||
(window.matchMedia(‘(prefers-color-scheme: dark)’).matches ? ‘dark’ : ‘light’);

// Modify player theme based on localStorage value.
let options = {“autoplay”:false,”hideControls”:null,”language”:”en-us”,”loop”:false,”partnerName”:”cloud-blogs”,”poster”:”https://azure.microsoft.com/en-us/blog/wp-content/uploads/2026/03/the-shift-1.jpg”,”title”:””,”sources”:[{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/3233302-Shift-Season2-Trailer-0x1080-6439k”,”type”:”video/mp4″,”quality”:”HQ”},{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/3233302-Shift-Season2-Trailer-0x720-3266k”,”type”:”video/mp4″,”quality”:”HD”},{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/3233302-Shift-Season2-Trailer-0x540-2160k”,”type”:”video/mp4″,”quality”:”SD”},{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/3233302-Shift-Season2-Trailer-0x360-958k”,”type”:”video/mp4″,”quality”:”LO”}],”ccFiles”:[{“url”:”https://azure.microsoft.com/en-us/blog/wp-json/msxcm/v1/get-captions?url=https%3A%2F%2Fwww.microsoft.com%2Fcontent%2Fdam%2Fmicrosoft%2Fbade%2Fvideos%2Fproducts-and-services%2Fen-us%2Fazure%2F3233302-shift-season-2-trailer%2F3233302-Shift-Season-2-Trailer_cc_en-us.ttml”,”locale”:”en-us”,”ccType”:”TTML”}],”downloadableFiles”:[{“url”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/3233302-Shift-Season-2-Trailer_transcript_en-us”,”locale”:”en-us”,”mediaType”:”transcript”},{“url”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/3233302-Shift-Season-2-Trailer_audio_en-us”,”locale”:”en-us”,”mediaType”:”audio”}]};

if (currentTheme) {
options.playButtonTheme = currentTheme;
}

document.addEventListener(‘DOMContentLoaded’, () => {
ump(“ump-69b081f9c8544″, options);
});

Topics we’ll explore weekly

Are my agents hunting for data?

How do agents work together?

Wait, my agent needs a database?

Is context engineering the new RAG?

What senses do my agents need to act?

Is Postgres the wave of the future?

Should my IT team hire agents?

How do we draw agentic borders?

Agents don’t succeed in isolation. They depend on how your data is unified, how your cloud handles scale, how your applications orchestrate across systems, and ultimately, how this serves people. At Microsoft, we see agents as catalysts for innovation across your entire environment, performing best when layers of the stack work together. That’s where the toughest challenges for technical teams emerge: observability, governance, security, optimization, and quality. It’s a team sport.

Your data strategy determines what your agents can reason over. Your cloud foundation determines what you can do reliably. Your agents and AI app experiences deliver business outcomes. Our colleagues and friends featured on The Shift are solving for these interdependencies. And what they all have in common is conviction that none of this works in pieces.

Our first episode, “Are my agents hunting for data?” drops tomorrow. We’ll sit with Ronald Chang, Dipti Borkar, Josh Caplan, and Cillian Mitchell from the Microsoft Fabric and Microsoft OneLake teams to cover why data preparation is essential to fueling agents with knowledge. And it’s perfect timing with Microsoft Fabric Community Conference next week in Atlanta. I hope you’ll join us to keep this conversation going.

Subscribe today on YouTube, Spotify, Apple Podcasts, Amazon Music, RSS.com, or wherever you listen and learn.

The agentic shift starts here
Follow us on YouTube to get the latest episodes.

Subscribe to the podcast

The post Unpacking your top questions on agentic AI: The Shift podcast appeared first on Microsoft Azure Blog.
Quelle: Azure

Azure IaaS series: Explore new resources for building a stronger, more efficient infrastructure

Why a modern cloud infrastructure foundation is critical to your business

Infrastructure has always been foundational to running business-critical cloud workloads; but today, it has become a strategic driver of innovation, resilience, and growth. As organizations accelerate digital transformation, infrastructure decisions increasingly shape how quickly teams can adopt AI, how reliably applications operate at global scale, and how effectively businesses respond to constant change.

To help customers navigate this shift, we’re introducing the Azure IaaS (Infrastructure as a Service) Resource Center: a centralized destination that brings together guidance, resources, demos, architectures, and best practices to support infrastructure design, optimization, and operations across compute, storage, and networking.

How does IaaS provide scalable cloud infrastructure?

AI adoption is accelerating faster than most organizations can operationalize it, with the pace and complexity of this shift becoming unprecedented. Applications are becoming more distributed and data intensive, while expectations for performance, availability, and security continue to rise. At the same time, leaders face growing pressure to optimize costs and ensure infrastructure investments align to tangible business outcomes.

These pressures are showing up in real, day-to-day infrastructure decisions:

Designing for continuity as environments grow more distributed and interdependent.

Strengthening security and compliance in an increasingly sophisticated threat landscape.

Achieving the performance required for data-intensive, latency-sensitive, and AI-driven workloads.

Keeping infrastructure flexible as workload patterns evolve and business priorities change.

Optimizing spend while ensuring infrastructure decisions are aligned with actual workload requirements.

This is exactly where a more intentional infrastructure strategy becomes critical. What has changed is not just the scale of infrastructure, but the need for system-level design across compute, storage, and networking. Infrastructure can no longer be optimized in isolation or managed reactively. It must operate as a cohesive platform, where performance, resiliency, security, scalability, and cost efficiency reinforce one another.

Azure IaaS has been designed for this reality, providing the foundation to run your most important cloud workloads today, while giving you the flexibility to adapt as needs evolve. To help organizations navigate this shift with clarity and confidence, the new Azure IaaS Resource Center offers a centralized destination to explore the guidance, resources, demos, architectures, and best practices needed to design, optimize, and operate infrastructure with confidence across every layer of the stack.

const currentTheme =
localStorage.getItem(‘msxcmCurrentTheme’) ||
(window.matchMedia(‘(prefers-color-scheme: dark)’).matches ? ‘dark’ : ‘light’);

// Modify player theme based on localStorage value.
let options = {“autoplay”:false,”hideControls”:null,”language”:”en-us”,”loop”:false,”partnerName”:”cloud-blogs”,”poster”:”https://cdn-dynmedia-1.microsoft.com/is/image/microsoftcorp/1620950-AzureIaaSBuild_tbmnl_en-us?wid=1280″,”title”:””,”sources”:[{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/1620950-AzureIaaSBuild-0x1080-6439k”,”type”:”video/mp4″,”quality”:”HQ”},{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/1620950-AzureIaaSBuild-0x720-3266k”,”type”:”video/mp4″,”quality”:”HD”},{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/1620950-AzureIaaSBuild-0x540-2160k”,”type”:”video/mp4″,”quality”:”SD”},{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/1620950-AzureIaaSBuild-0x360-958k”,”type”:”video/mp4″,”quality”:”LO”}],”ccFiles”:[{“url”:”https://azure.microsoft.com/en-us/blog/wp-json/msxcm/v1/get-captions?url=https%3A%2F%2Fwww.microsoft.com%2Fcontent%2Fdam%2Fmicrosoft%2Fbade%2Fvideos%2Fproducts-and-services%2Fen-us%2Fazure%2F1620950-azureiaasbuild%2F1620950-AzureIaaSBuild_cc_en-us.ttml”,”locale”:”en-us”,”ccType”:”TTML”}],”downloadableFiles”:[{“url”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/1620950-AzureIaaSBuild_transcript_en-us”,”locale”:”en-us”,”mediaType”:”transcript”},{“url”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/1620950-AzureIaaSBuild_audio_en-us”,”locale”:”en-us”,”mediaType”:”audio”}]};

if (currentTheme) {
options.playButtonTheme = currentTheme;
}

document.addEventListener(‘DOMContentLoaded’, () => {
ump(“ump-69a9cbdcd7a57″, options);
});

A modern infrastructure platform engineered for performance, security, and global scale

Azure IaaS brings together a comprehensive portfolio of compute, storage, and networking services to support a wide range of workloads, from: line-of-business applications and databases to analytics platforms, AI training clusters, and global consumer applications.

Built with a system-level approach, Azure IaaS unifies specialized hardware, intelligent software, high-capacity networking, and platform orchestration to deliver consistent performance, strong security protections, and flexible scaling. Backed by more than 70 regions worldwide, a private global fiber backbone, hardware acceleration, integrated resiliency, and multilayer security, Azure provides an infrastructure foundation ready for modern and future business demands.

Resilient by design to help keep your business running

Azure’s infrastructure is built from the ground up for resilience, ensuring applications remain available even when the unexpected occurs. With a broad portfolio of infrastructure options spanning zonal redundancy, regional redundancy, and globally distributed architectures, organizations can architect for continuity at every layer.

Azure’s compute, storage, and networking platforms are engineered to withstand failure through intelligent load balancing, fast failover mechanisms, and integrated data protection. This resilient foundation empowers organizations to operate with confidence, whether running mission-critical systems that demand continuous uptime or scaling AI-driven applications that cannot tolerate disruption.

By combining proactive fault isolation, automated recovery, and multilayer redundancy, Azure IaaS helps organizations maintain operations through outages, recover rapidly, and safeguard the business against uncertainty.

With Azure, resilience isn’t an addon; it’s the architecture that helps your infrastructure keep pace with your most ambitious goals.

High-performance Azure IaaS for your most demanding workloads—from databases to AI clusters

With a comprehensive portfolio of Azure Virtual Machine series—including memory-optimized, compute-optimized, GPU-accelerated, and storage-optimized options—customers can match infrastructure precisely to their workload needs, whether running mission-critical databases or training advanced AI models. The latest VM families leverage cutting-edge processors and high-speed networking, enabling ultra-low latency and massive throughput for data-intensive and AI-driven applications. This flexibility empowers organizations to match their infrastructure choices to their specific workload needs, harnessing the same platform for both everyday business operations and the most demanding AI workloads. As a result, Azure IaaS provides the foundation for innovation to help ensure your infrastructure keeps pace with your boldest goals.

Built-in security and compliance on Azure IaaS to help reduce risk

Security on Azure IaaS is a top priority; engineered into the platform across compute, storage, and networking. From the underlying hardware to the workloads it supports, Azure applies a defense-in-depth approach designed to protect infrastructure as threats continue to evolve.

At the foundation, Azure security includes secure supply chain practices, a rigorous secure development lifecycle (SDL), encryption, and identity and access management with Microsoft Entra ID.

Networking security helps reduce exposure through isolation, segmentation, and private connectivity, using virtual networks, Network Security Groups, and Private Link to limit public access. Services such as Azure Firewall and DDoS Protection add protection and control at scale.

Storage security enforces encryption by default, provides identity-based access controls, and includes safeguards such as soft delete, versioning, and immutability to reduce the risk of loss or tampering.

Compute security is rooted in hardware-based trust, starting with server-level secure boot and attestation, VM-level capabilities like Trusted Launch, secure VM boot, and a virtual Trusted Platform Module, and Azure confidential computing to help protect workloads and sensitive data in use.

Together, these integrated protections help organizations reduce risk, meet compliance requirements, and run critical infrastructure securely—without slowing innovation.

Scale infrastructure with flexibility to support changing workload needs

Modern workloads place uneven and evolving demands on infrastructure. Capacity must expand quickly, scale independently across layers, and extend globally.

Azure IaaS enables this flexibility by providing extensive solutions to scale compute, storage, and networking independently based on actual workload requirements. Teams can compute vertically by increasing VM sizes and performance levels, or horizontally by intelligently distributing workloads across multiple VM types, availability zones, and regions. Storage capacity and performance can be adjusted separately to support data growth and throughput needs, while high-capacity networking enables low-latency connectivity across distributed environments.

With more than 70 regions worldwide, Azure IaaS provides a variety of solutions that supports geographic expansion and proximity to users and data. Azure IaaS continues to innovate on deployment and capacity management solutions that provide users with increased scalability and decreased overhead. Global networking and region-to-region connectivity make it possible to scale applications while maintaining consistent performance and availability.

Together, elastic infrastructure, global reach, and adaptive architectural patterns help organizations expand capacity, respond to demand shifts, and support growth.

Build a cost-efficient cloud infrastructure strategy with Azure IaaS

Cost optimization in the cloud is about reducing spend while making informed infrastructure decisions that balance efficiency, performance, and business value. As workloads grow more complex and data-intensive, organizations are looking not only to lower costs, but to ensure every dollar invested in infrastructure delivers measurable impact.

Azure IaaS is designed to support this balance. It gives organizations the flexibility to optimize costs based on real workload requirements; whether that means right-sizing compute resources, aligning storage performance to actual usage, or selecting networking options that meet throughput needs without overprovisioning. By matching infrastructure capabilities to demand, teams can reduce unnecessary spend, while maintaining the performance and reliability their applications require.

Optimal cost efficiency on Azure is not a one-time exercise either. Built-in tooling and guidance help teams continuously evaluate usage patterns, identify inefficiencies, and adapt as workloads evolve. Flexible pricing options such as reservations and savings plans enable predictable cost control for steady-state workloads, while elastic scaling models support dynamic environments where demand fluctuates.

Azure IaaS also helps organizations optimize costs by reducing operational overhead. Managed services, automation, and integrated monitoring simplify infrastructure management, allowing teams to focus on improving utilization and performance rather than managing complexity. For organizations modernizing or migrating workloads, Azure provides purpose-built tools that help transition data and applications efficiently; creating opportunities that reduce long-term costs, while improving operational consistency.

Whether supporting core business systems, scaling global applications, or enabling AI innovation, with Azure IaaS you can reduce costs, improve price-performance, and continuously optimize infrastructure investments. Cost efficiency becomes not a constraint on innovation, but a foundation that enables it.

Your infrastructure for the AI era starts with Azure

AI is changing the demands placed on infrastructure. Teams are moving beyond experimentation to operationalizing AI across the business: training models, running inference at scale, and integrating AI into line-of-business applications and decision workflows. That shift requires more than raw computing power. It depends on an infrastructure platform that can deliver the right combination of performance, resiliency, security, scalability, and cost efficiency—together.

Azure IaaS is designed to support the full spectrum of AI workloads, helping organizations bring AI workloads closer to users and data—reducing latency and improving responsiveness. With integrated resiliency capabilities and multi-layered security, Azure supports the continuity and protection required for business-critical AI scenarios. And with flexible infrastructure choices and optimization models, organizations can scale AI responsibly while maintaining control over spend.

As AI requirements evolve quickly, the ability to make infrastructure decisions with clarity matters. The Azure IaaS Resource Center can help you navigate those decisions to connect the guidance, best practices, and practical resources needed to move from planning to production with confidence.

Build confidently, run efficiently, and innovate boldly with Azure IaaS

Whether you’re modernizing mission-critical systems, supporting global applications, optimizing hybrid and multi-cloud environments, or preparing your organization for AI innovation, Azure IaaS provides the trusted infrastructure platform to help you move forward—without trading off performance, resiliency, security, scalability, or cost efficiency.

The Azure IaaS Resource Center is your central destination to explore best practices, learn from experts, and find the right guidance for every stage of your infrastructure journey across compute, storage, and networking.

Build in the cloud with Azure
Visit the Azure IaaS Resource Center to start building a stronger, more efficient infrastructure today.

Get started with Azure

The post Azure IaaS series: Explore new resources for building a stronger, more efficient infrastructure appeared first on Microsoft Azure Blog.
Quelle: Azure

Instant access incremental snapshots: Restore without waiting

Today, we’re excited to introduce instant access support for incremental snapshots of Premium SSD v2 (Pv2) and Ultra Disk, delivering an industry-leading snapshot experience where creation, disk restore, and production-ready performance all happen instantly.

Instant access snapshots of Pv2 and Ultra Disk are usable immediately upon creation, allowing you to restore new disks without waiting for snapshots to be ready. Newly restored disks also deliver near-full performance from the start, allowing applications to resume operations immediately.

For mission-critical workloads, speed matters—especially when every minute of downtime or degraded performance has real business impact. Whether you’re rolling out critical application updates, recovering unexpected issues, scaling stateful workloads, or refreshing environments for development and testing, snapshots play a central role in keeping your business moving forward.

Standard incremental snapshots have long provided a cost-effective method to protect data and create point-in-time disk copies for these workflows. However, this came with tradeoffs: after snapshot creation, data must be fully copied before snapshots can be used to restore disks, and disks must fully hydrate to reach full performance.

With instant access snapshots, that wait is gone.

As soon as snapshots are created, you can restore new disks from them with near-full performance from the start. This means faster recovery, faster scaling, and faster iteration—redefining what customers should expect from snapshot operations in cloud.

See instant access documentation

Instant access snapshot in action

When creating incremental snapshots of Pv2 and Ultra Disk, you can opt in to enable instant access using the existing snapshot API. Once enabled, snapshots enter the Instant Access state at creation and can be used to create new disks right away with fast restore performance.

Instant access snapshots come with the following benefits:

Instant availability: Instant access snapshots are available to restore new disks immediately upon creation, eliminating the wait time for background data copy to snapshot in Standard storage.

Fast restore: Disks restored from instant access snapshots hydrate rapidly and deliver near-full performance with single-digit millisecond read latencies and sub-millisecond write latencies from the start.

Incremental storage: Instant access snapshots store only the incremental changes made to the source disk after snapshot creation, avoiding the need for a full base snapshot to keep your storage costs low.

Cross-zonal restore: Like incremental snapshots in Standard storage, instant access snapshots support restoring disks into a different availability zone within the same region, providing flexibility for deployment and recovery.

These benefits make instant access snapshots best suited for time-sensitive scenarios where immediate availability and performance are essential.

Common use cases include:

Fast rollback: Before deploying major software upgrades, you can use instant access snapshots to create an immediate recovery point. If issues occur such as accidental deletes, you can restore disks from snapshots to quickly rollback changes, maintain business continuity, and minimize downtime.

Uninterrupted maintenance: Instant access snapshots let you begin maintenance immediately after snapshot creation—without waiting for them to become ready. This reduces application downtime and keeps large systems within tight maintenance windows.

Rapid scale-out: You can scale stateful applications by leveraging instant access snapshots to create multiple disk copies of your primary instance in seconds, such as adding secondary SQL Server replicas. You can also create copies across availability zones within the same region quickly to achieve high availability and resiliency.

Refresh secondary environments: Instant access snapshots let you copy production disks and spin up secondary environments for development and testing in seconds. This ensures your test environment is always in sync with production and delivers high performance right away.

Here’s what instant access snapshots looks like in practice

For Azure Database Services such as PostgreSQL, instant access snapshots are a core building block for backup, recovery, and scaling. Consider a SaaS provider running PostgreSQL for a customer-facing application. During peak business hours, read traffic spikes requiring rapid scale-out of read capacity. With instant access snapshots, the Azure Database Service can create snapshots from the existing instance and quickly create new disks and instances with zero wait time. This allows new read replicas to be brought online much faster than traditional snapshot workflows, immediately serving read traffic with low latency and allowing the application to absorb traffic surges on demand.

The customer perspective below demonstrates how instant access snapshots transform their day-to-day operations:

For Providence, instant access snapshots have transformed how we refresh our Epic training and support environments. Instead of waiting on long background copy processes for snapshot creation and disk hydration, our teams can create fully usable disks immediately for the day’s training and support workflows. The ability to refresh downstream environments quickly has been a massive boost to our operational efficiency and has significantly improved the experience for our end users.
— Chris Calas, Senior Principal Service Engineer, Providence

At Microsoft Ignite 2025, instant access support for incremental snapshots is highlighted as a key innovation by Azure Storage product leaders, underscoring Azure’s focus on mission-critical workloads.

const currentTheme =
localStorage.getItem(‘msxcmCurrentTheme’) ||
(window.matchMedia(‘(prefers-color-scheme: dark)’).matches ? ‘dark’ : ‘light’);

// Modify player theme based on localStorage value.
let options = {“autoplay”:false,”hideControls”:null,”language”:”en-us”,”loop”:false,”partnerName”:”cloud-blogs”,”poster”:”https://azure.microsoft.com/en-us/blog/wp-content/uploads/2026/03/Ultra-Disk-Video.jpg”,”title”:””,”sources”:[{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/957600-AzureStorageinnovationswiththenextgeneration-0x1080-6439k”,”type”:”video/mp4″,”quality”:”HQ”},{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/957600-AzureStorageinnovationswiththenextgeneration-0x720-3266k”,”type”:”video/mp4″,”quality”:”HD”},{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/957600-AzureStorageinnovationswiththenextgeneration-0x540-2160k”,”type”:”video/mp4″,”quality”:”SD”},{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/957600-AzureStorageinnovationswiththenextgeneration-0x360-958k”,”type”:”video/mp4″,”quality”:”LO”}],”ccFiles”:[{“url”:”https://azure.microsoft.com/en-us/blog/wp-json/msxcm/v1/get-captions?url=https%3A%2F%2Fwww.microsoft.com%2Fcontent%2Fdam%2Fmicrosoft%2Fbade%2Fvideos%2Fproducts-and-services%2Fen-us%2Fazure%2F957600-azurestorageinnovationswiththenextgeneration%2F957600-AzureStorageinnovationswiththenextgeneration_cc_en-us.ttml”,”locale”:”en-us”,”ccType”:”TTML”}],”downloadableFiles”:[{“url”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/957600-AzureStorageinnovationswiththenextgeneration_transcript_en-us”,”locale”:”en-us”,”mediaType”:”transcript”},{“url”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/957600-AzureStorageinnovationswiththenextgeneration_audio_en-us”,”locale”:”en-us”,”mediaType”:”audio”}]};

if (currentTheme) {
options.playButtonTheme = currentTheme;
}

document.addEventListener(‘DOMContentLoaded’, () => {
ump(“ump-69a74a6293c14″, options);
});

How instant access snapshots improve disk restore performance

Standard incremental snapshots for Pv2 and Ultra Disk are designed to optimize cost efficiency and durability by storing point-in-time data in Standard Zone-Redundant Storage (ZRS). This approach provides strong resilience and is well suited for long-term backups and protection against zonal disasters. Once snapshot data has been fully copied into Standard ZRS, it becomes available for restoring new disks. A disk restored from a Standard storage snapshot can be used immediately, but until the background data copy to the disk completes, reads are served from Standard storage, resulting in slower performance and a longer copy time.

Incremental snapshots with instant access enabled take a different approach to optimize instant availability and restore performance. With instant access snapshots, point-in-time data captured remains in the same high-performance storage location as the source disks for the duration specified. As a result, snapshots are available to restore new disks immediately upon creation, without waiting for snapshot background data copy to be completed. Newly restored disks from instant access snapshots benefit from high hydration bandwidth and low latency from the start, as data is served directly from instant access snapshots.

When snapshots are created with instant access enabled, the system begins copying the data to Standard storage in parallel. This ensures that once the instant access duration expires, a fully hardened Standard ZRS snapshot is automatically available, without requiring any additional action.

This architecture enables faster recovery, quicker environment refreshes, and more predictable performance for time-sensitive operations.

How to create instant access snapshots

Instant access snapshots are not a separate snapshot resource class to manage. They are incremental snapshots that temporarily enter Instant Access state for the specified duration. When this duration expires, snapshots automatically transition out of the Instant Access state and continue as Standard ZRS snapshots for better reliability and long-term retention.

If you are already familiar with creating incremental snapshots with existing Azure Snapshot API, enabling instant access support will feel completely natural. The same commands and API endpoints apply; the only difference is a single additional parameter that defines how long the snapshots remain in the Instant Access state. The Azure CLI example below illustrates how to enable instant access when creating incremental snapshots.

Here’s how you create incremental snapshots using Azure CLI today:

< az snapshot create –resource-group <rg-name> –name <snapshot-name> –source <disk-id> –incremental true >

To enable instant access, simply include the InstantAccessDurationMins parameter:

< az snapshot create –resource-group <rg-name> –name <snapshot-name> –source <disk-id> –incremental true   –InstantAccessDurationMins 300 >

The technical demo video below demonstrates how to create instant access snapshots and restore disks in Azure Portal. You will also learn how to easily integrate this capability into your existing snapshot workflow.

const currentTheme =
localStorage.getItem(‘msxcmCurrentTheme’) ||
(window.matchMedia(‘(prefers-color-scheme: dark)’).matches ? ‘dark’ : ‘light’);

// Modify player theme based on localStorage value.
let options = {“autoplay”:false,”hideControls”:null,”language”:”en-us”,”loop”:false,”partnerName”:”cloud-blogs”,”poster”:”http://cdn-dynmedia-1.microsoft.com/is/image/microsoftcorp/1593700-InstantAccessSnapshots_tbmnl_en-us?wid=1280″,”title”:””,”sources”:[{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/1593700-InstantAccessSnapshots-0x1080-6439k”,”type”:”video/mp4″,”quality”:”HQ”},{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/1593700-InstantAccessSnapshots-0x720-3266k”,”type”:”video/mp4″,”quality”:”HD”},{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/1593700-InstantAccessSnapshots-0x540-2160k”,”type”:”video/mp4″,”quality”:”SD”},{“src”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/1593700-InstantAccessSnapshots-0x360-958k”,”type”:”video/mp4″,”quality”:”LO”}],”ccFiles”:[{“url”:”https://azure.microsoft.com/en-us/blog/wp-json/msxcm/v1/get-captions?url=https%3A%2F%2Fwww.microsoft.com%2Fcontent%2Fdam%2Fmicrosoft%2Fbade%2Fvideos%2Fproducts-and-services%2Fen-us%2Fazure%2F1593700-instantaccesssnapshots%2F1593700-InstantAccessSnapshots_cc_en-us.ttml”,”locale”:”en-us”,”ccType”:”TTML”}],”downloadableFiles”:[{“url”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/1593700-InstantAccessSnapshots_transcript_en-us”,”locale”:”en-us”,”mediaType”:”transcript”},{“url”:”https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/1593700-InstantAccessSnapshots_audio_en-us”,”locale”:”en-us”,”mediaType”:”audio”}]};

if (currentTheme) {
options.playButtonTheme = currentTheme;
}

document.addEventListener(‘DOMContentLoaded’, () => {
ump(“ump-69a74a6295694″, options);
});

Availability and pricing

Instant access support for incremental snapshots of Premium SSD v2 and Ultra Disk is available in all Azure regions where these disk types are supported, ensuring the capability is available wherever supported mission-critical workloads run.

Instant access support uses a simple, usage-based billing model that charges only for the additional storage capacity consumed and for each restore operation. This approach keeps costs closely aligned with your actual usage, making cloud spend easier to understand and forecast. For full pricing details, see Managed Disk Pricing.

Getting started

As organizations continue to run increasingly time-sensitive mission critical workloads in Azure, instant access support for incremental snapshots of Premium SSD v2 and Ultra Disk provides a practical way to improve recovery time, streamline operations, and scale with confidence.

Explore instant access snapshots documentation
Learn everything you need to know for Azure managed disks.

Explore more

The post Instant access incremental snapshots: Restore without waiting appeared first on Microsoft Azure Blog.
Quelle: Azure

Unlocking document understanding with Mistral Document AI in Microsoft Foundry

Enterprises today face a familiar yet formidable challenge: mountains of documents -contracts, invoices, reports, forms – remain locked in unstructured formats. Traditional OCR (optical character recognition) captures text, but often struggles with context, layout complexity, or multilingual content. The result? Slow workflows, error-prone manual reviews, and missed insights.

Enter mistral-document-ai-2512 in Microsoft Foundry. This new model brings together high-end OCR using mistral-ocr-2512 and intelligent document understanding using mistral-small-2506 to turn unstructured documents into actionable data. It doesn’t just “read” pages – it understands them: multi-column layouts, handwritten annotations, tables with merging cells, multilingual content-all processed with enterprise-grade speed and precision.

In this blog, we’ll explore what Mistral Document AI 2512 is, why it matters, how it stacks up, and the business impact it promises, especially when paired with solution accelerators like ARGUS.

Meet Mistral Document AI

Mistral Document AI is an enterprise-grade document understanding model, offered via Microsoft Foundry. It’s built to convert both physical (scans, photos) and digital (PDFs, DOCX) documents into highly structured, machine-readable outputs. Key features include:

Top-tier accuracy: According to benchmarks, Mistral’s OCR 2512 stacks display significantly higher accuracy than many alternatives, especially on scanned documents and complex layouts. For example, in comparisons it achieved ~95.9 % “overall” vs ~89-91 % for other platformsGlobal / multilingual reach: In language-by-language tests (Russian, French, German, Spanish, Chinese, etc), Mistral’s error-rate/fuzzy-match metrics reached 99 %+ in many casesLayout & context awareness: It’s built to not just extract linear text but understand multi-column layouts, tables, charts, images, handwritten input and moreStructured output functionality: The model supports structured extraction (JSON), markup (Markdown with interleaved images), preserving document structure for downstream systemsEnterprise-ready deployment: With availability via Microsoft Foundry and support for private/secure inference, the model is geared for regulated industries and high-volume workflowsPutting it another way: where traditional OCR stops at “here’s the raw text on page 7”, Mistral DocumentAI 2512 can say “here’s the vendor invoice, here are line-items, here’s the total, here’s the signature block, and here’s the part that was handwritten”, ready to plug into downstream systems.

Business Impact & Industry examples

Mistral Document AI isn’t just another OCR tool; it’s a strategic enabler that turns document-heavy operations into intelligent, automated workflows. The business value comes down to four key advantages:

Speed and efficiency: Automating document understanding eliminates manual reviews and retyping. Tasks that took days can be done in minutes, accelerating core business processesAccuracy and consistency: With 99 %+ recognition accuracy and deep layout understanding, Mistral delivers cleaner data and fewer downstream errors – essential in compliance-critical or analytics-driven operationsCost and productivity gains: Reducing manual extraction frees teams for higher-value work, cutting operational costs while increasing output per employeeScalability and adaptability: Cloud-native performance allows organizations to scale document processing instantly during peak loads, across multiple languages and formats, without sacrificing qualityOverall, mistral-document-ai-2512 excels where consistency and quality are critical.

Industry and Use CasesIn regulated industries or big-data scenarios, even a small improvement in accuracy or speed can translate into substantial business gains. Its benchmarks indicate not just incremental progress, but a major step forward – giving enterprises a powerful new engine for their document workflows.

Here’s where that impact becomes tangible:

Financial services: Banks and insurers handle vast document volumes – loan applications, KYC forms, and claims reports – where data integrity and auditability are non-negotiable. Mistral automates extraction, classification, and clause identification across diverse formats, improving turnaround time and compliance accuracy while reducing manual handling costs

Healthcare & life sciences: Clinical records, lab results, and insurance claims often combine handwritten, tabular, and multi-language content. Mistral’s layout awareness and multilingual support ensure clean, structured datasets for downstream analytics and regulatory submissions

Manufacturing & logistics: From quality certificates to shipping manifests, Mistral streamlines the flow of operational documents. It can extract production parameters, vendor data, and timestamps at scale – building a unified, queryable data layer that supports supply chain traceability

Legal & public sector: Legal teams and agencies depend on consistency and transparency. Mistral helps index, summarise, and validate contracts or permits with full structural fidelity – dramatically cutting review cycles while maintaining evidential quality

Retail & consumer goods: Retailers process supplier invoices, product specifications, and marketing briefs from global partners. With Mistral’s multilingual precision and structure preservation, global document flows become searchable and analytics-ready

Across these industries, the result is the same: cleaner data, faster throughput, and fewer human errors – the foundation for more reliable decisions and more agile operations.

Pricing

Argus – A ready-to-implement accelerator to start using Mistral Document AI

To spin up a solution faster, one can leverage solution accelerators such as ARGUS (open-source repository available on GitHub).ARGUS serves as a full-pipeline implementation: from document ingestion, OCR/extraction (via Mistral Document AI), to downstream processing and structured output. It shows how to deploy end-to-end, integrate with storage, preprocess documents, handle large-scale batches, output JSON schemas, and integrate into existing business workflows.

Mistral Document AI Integration

ARGUS now offers flexible OCR provider selection with Mistral Document AI as one of the several options. This enhancement gives you the freedom to choose the best OCR engine for your specific document processing needs.

Key Features:

Dual Provider Support: Toggle between Azure Document Intelligence (default) and Mistral Document AIRuntime Switching: Change OCR providers on-the-fly through the Settings UI without redeploymentSimple Configuration: Set up Mistral via environment variables (OCR_PROVIDER, MISTRAL_DOC_AI_ENDPOINT, MISTRAL_DOC_AI_KEY) or the web interfaceSeamless Integration: Both providers expose the same interface, ensuring consistent behavior across your document processing pipelineWhy This Matters:

Different OCR engines excel at processing different document content. Azure Document Intelligence offers enterprise-grade form and table recognition, while Mistral Document AI 2512, in addition, enables extraction to structured JSON with customizable schemas, document classification, and image processing—including text, charts, and signatures. It can convert charts into tables, extract fine print from figures, and even define custom image types for specialized workflows. Now you can select the optimal provider for each use case.

In effect, instead of building from scratch, ARGUS gives you the legs to run: pipeline orchestration, ingestion, error-handling, schema-mapping, output integration-all wired to Mistral’s engine. This significantly accelerates time-to-value and reduces risk for enterprise adopters.

Getting Started:

Navigate to the ARGUS frontend interface (Streamlit app) and click on the Settings tab. In the OCR Provider Configuration section, select your preferred provider. If using Mistral, enter your endpoint URL, API key, and model name. Click Update OCR Provider to apply changes immediately—no restart required. All new document processing will use your selected OCR engine.

If your organization is looking to unlock document intelligence, here’s a structured path:

Explore Mistral Document AI via Microsoft Foundry: Browse the model card, review endpoint specs, try sample documents to test accuracy and extraction structureDeploy and Pilot with ARGUS: Use the GitHub repo to spin up an end-to-end pipeline on a small workload (e.g., a batch of invoices or contracts) and compare manual vs AI-driven throughput and error-ratesDefine business value metrics: Track processing time, error rate, manual hours saved, and downstream impact (faster decision cycles, fewer reworks).Scale and govern: Once pilot proves value, expand into multiple document types, languages, geographies – and ensure governance (data handling, compliance, model-monitoring)Embed continuous improvement: As usage grows, feed back learnings, tune schema definitions, refine extraction rules, and extend into QA, insights or analytics layers

Conclusion

In today’s data-rich but document-heavy environment, the ability to truly understand documents (and not just digitize them) is becoming a strategic imperative. Mistral Document AI represents a next-generation shift: accurate, layout-aware, multilingual, structured. When paired with accelerators like ARGUS, enterprises can move from manual bottlenecks to streamlined, insight-rich document workflows.

If you’re thinking about unlocking the value buried in your documents-be it invoices, contracts, forms or reports, now is the time. With mistral-document-ai-2512, what used to be a cost-center is now a potential performance lever.

Ready to get started? Explore the model, and let your documents begin talking back.
The post Unlocking document understanding with Mistral Document AI in Microsoft Foundry appeared first on Microsoft Azure Blog.
Quelle: Azure

Microsoft Sovereign Cloud adds governance, productivity, and support for large AI models securely running even when completely disconnected

As digital sovereignty becomes a strategic requirement, organizations are rethinking how they deploy critical infrastructure and AI capabilities under tighter regulatory expectations and higher risk conditions. Microsoft’s approach to sovereignty is grounded in enabling enterprises, public sectors, and regulated industries to participate in the digital economy securely, independently, and on their own terms. The Microsoft Sovereign Cloud brings together productivity, security, and cloud workloads to span both public and private environments so organizations can choose the right level of control, capability, and compliance. Customers can choose the right control posture for each workload, through a continuum of sovereign options protecting against fragmenting their architecture or increasing operational risk. Trust is built on confidence: confidence that data stays protected, controls are enforceable, and operations can continue under real-world conditions.

To support these confidential environments, Microsoft offers full stack capabilities that support customers across connected, intermittently connected, and fully disconnected modes. Today’s expansion of capabilities includes three major updates:

Azure Local disconnected operations (now available) – Organizations can now run mission-critical infrastructure with Azure governance and policy control, with no cloud connectivity, optimizing continuity for sovereign, classified or isolated environments.

Microsoft 365 Local disconnected (now available) – Core productivity workloads, Exchange Server, SharePoint Server, and Skype for Business Server, can run fully inside the customer’s sovereign operational boundary on Azure Local, keeping teams productive even when disconnected from the cloud.

Foundry Local adds large model and modern infrastructure capabilities – Organizations can now bring large AI models into fully disconnected, sovereign environments with Foundry Local. Using modern infrastructure from partners like NVIDIA, customers with sovereign needs will now be able to run multimodal models locally on their own hardware, inside strict sovereign boundaries enabling powerful, local AI inferencing in fully disconnected environments.

This delivers a truly localized full stack experience built on Azure Local infrastructure and Microsoft 365 Local workloads, designed to stay resilient across any connectivity condition, with large models being part of Foundry Local extending the stack to run advanced multimodal models locally, securely, even when fully disconnected. Customers can now help maintain uninterrupted operations, keep mission critical workloads protected, and apply consistent governance and policy enforcement, while keeping data, identities, and operations within their sovereign boundaries.

Azure Local runs critical infrastructure locally, even when disconnected

For workloads with specialized requirements, Azure Local provides the on-premises foundation with consistent Azure governance and policy controls. With Azure Local disconnected operations, management, policy, and workload execution stay within the customer-operated environments, so services continue running securely even when environments must be isolated or connectivity is not available. Using familiar Azure experiences and consistent policies, organizations can deploy and govern workloads locally without depending on continuous connection to public cloud services. Azure Local is designed to scale with mission-critical needs from smaller deployments to larger footprints that support data-intensive and AI-driven workloads. Customers can start fast, expand over time, and maintain a unified operational model, all within their sovereign boundary.

Operating in disconnected environments surfaces constraints that go beyond traditional cloud assumptions: external dependencies may be unacceptable, connectivity may be intentionally restricted, and operational continuity is a business imperative.

“The availability of Azure Local disconnected operations represents a breakthrough for organizations that need control over their data without sacrificing the power of the Microsoft Cloud. For Luxembourg, where digital sovereignty is not just a principle but a strategic necessity, this model offers the resilience, autonomy, and trust our market expects. By combining Microsoft’s technological leadership with Proximus NXT’s sovereign cloud expertise, we are enabling our customers to innovate confidently – even in fully disconnected mode,”said Gerard Hoffmann, CEO Proximus Luxembourg.

Microsoft 365 Local keeps productivity and collaboration available in fully disconnected environments

As sovereign environments move into disconnected environments, keeping people productive becomes just as critical as keeping infrastructure online. Building on more than a decade of delivering and supporting these services, Microsoft 365 Local disconnected brings that continuity to the productivity layer, delivering Microsoft’s core server workloads—Exchange Server, SharePoint Server,and Skype for Business Server supported through at least 2035—directly into the customer’s sovereign private cloud.

With Microsoft 365 Local, teams can communicate, share information, and collaborate securely within the same controlled boundary as their infrastructure and AI workloads. Everything runs locally, under customer-owned policies, with full control of data resiliency, access, and compliance. By operating with Azure-consistent management and governance, customers get the productivity experience they rely on, designed to stay resilient and secure even when offline.

Bringing large models and modern infrastructure to Foundry Local

With the availability of larger models and modern infrastructure as part of the Foundry Local portfolio, Microsoft is enabling customers with highly secure environments the ability to run multimodal, large models directly inside their sovereign private cloud environments. This brings the richness of Microsoft’s enterprise AI capabilities to on-premises systems, complete with local inferencing and APIs that operate completely within customer-controlled data boundaries.

Expanding beyond small models, the integration of Foundry Local with Azure Local is specifically designed to support large-scale models utilizing the latest GPUs from partners such as NVIDIA. Microsoft will provide comprehensive support for deployments, updates, and operational health. Even as inferencing demands increase overtime, customers retain complete control over their data and hardware.

Choice and control without added complexity

Customers facing strict sovereignty and regulatory requirements are clear that a fully disconnected sovereign private cloud is a key business need. Microsoft Sovereign Private Cloud is designed to meet these needs head-on, enabling secure, compliant operations even in environments with no external connectivity. At the same time, we recognize that disconnected environments are not one-size-fits-all; some customers operate across connected, hybrid, and disconnected modes based on mission, risk, and regulation. Our approach helps customers to meet strict sovereign requirements in fully disconnected scenarios without compromising simplicity, while retaining flexibility where connectivity is possible. Together, Azure Local disconnected operations, Microsoft 365 Local, and Foundry Local help organizations choose where workloads run and how environments are managed, while standardizing governance and operational practices across connected and disconnected deployments.

Next steps

Get started

Azure Local disconnected operations and Microsoft 365 Local disconnected are now available worldwide, and large models on Foundry Local are available to qualified customers.

Explore the Microsoft Sovereign Cloud

Learn more about Azure Local disconnected operations

The post Microsoft Sovereign Cloud adds governance, productivity, and support for large AI models securely running even when completely disconnected appeared first on Microsoft Azure Blog.
Quelle: Azure

Azure reliability, resiliency, and recoverability: Build continuity by design

Modern cloud systems are expected to deliver more than uptime. Customers expect consistent performance, the ability to withstand disruption, and confidence that recovery is predictable and intentional.

In Azure, these expectations map the three distinct concepts: reliability, resiliency, and recoverability.

Explore technical methodologies with Azure Essentials

Reliability describes the degree to which a service or workload consistently performs at its intended service level within business-defined constraints and tradeoffs. Reliability is the outcome customers ultimately care about.

To achieve reliable outcomes, workloads are designed along two complementary dimensions. Resiliency is the ability to withstand faults and disruptive conditions such as infrastructure failures, zonal or regional outages, cyberattacks, or sudden change in load—and continue operating without customer-visible disruption. Recoverability is the ability to restore normal operations after disruption, returning the workload to a reliable state once resiliency limits are exceeded.

This blog anchors definitions and guidance to the Microsoft Cloud Adoption Framework, the Azure Well‑Architected Framework and the reliability guides for Azure services. Use the Reliability guides to confirm how each service behaves during faults, what protections are built in, and what you must configure and operate, so shared responsibility boundaries stay clear as workloads scale and during recovery scenarios.

Why this matters

When reliability, resiliency, and recoverability are used interchangeably, teams make the wrong design tradeoffs—over-investing in recovery when architectural resiliency is required, or assuming redundancy guarantees reliable outcomes. This post clarifies how these concepts differ, when each applies, and how they guide real design, migration, and incident-readiness decisions in Azure.

Industry perspective: Clarifying common confusion

Azure guidance treats reliability as the goal, achieved through deliberate resiliency and recoverability strategies. Resiliency describes workload behavior during disruption; recoverability describes restoring service after disruption.

Anchor principle: Reliability is the goal. Resiliency keeps you operational during disruption. Recoverability restores service when disruption exceeds design limits.

Part I — Reliability by design: Operating model and workload architecture

Reliable outcomes require alignment between organizational intent and workload architecture. Microsoft Cloud Adoption Framework helps organizations define governance, accountability, and continuity expectations that shape reliability priorities. Azure Well‑Architected Frameworktranslates those priorities into architectural principles, design patterns, and tradeoff guidance.

Part II — Reliability in practice: What you measure and operationalize

Reliability only matters if it is measured and sustained. Teams operationalize reliability by defining acceptable service levels, instrumenting steady-state behavior and customer experience, and validating assumptions with evidence.

Azure Monitor and Application Insights provide observability, while controlled fault testing (for example, with Azure Chaos Studio helps confirm designs behave as expected under stress.

Practical signals of “enough reliability” include meeting service levels for critical user flows, introducing changes safely, maintaining steady-state performance under expected load, and keeping deployment risk low through disciplined change practices.

Governance mechanisms such as Azure Policy, Azure landing zones, and Azure Verified Modules help apply these practices consistently as environments evolve.

The Reliability Maturity Model can help teams assess how consistently reliability practices are applied as workloads evolve, while remaining scoped to reliability practices rather than resiliency or recoverability architecture.

Part III — Resiliency in practice: From principle to staying operational

Resiliency by design is no longer a late-stage high-availability checklist. For mission-critical workloads, resiliency must be intentional, measurable, and continuously validated—built into how applications are designed, deployed, and operated.

Resiliency by design aims to keep systems operating through disruption wherever possible, not only recover after failures.

Resiliency is a lifecycle, not a feature

Effective practice shifts from isolated configurations to a repeatable lifecycle applied across workloads:

Start resilient—embed resiliency at design time using prescriptive architectures, secure-by-default configurations, and platform-native protections.  

Get resilient—assess existing applications, identify resiliency gaps, and remediate risks, prioritizing production mission-critical workloads. 

Stay resilient—continuously validate, monitor, and improve posture, ensuring configurations don’t drift and assumptions hold as scale, usage patterns, and threat models change.  

Withstanding disruption through architectural design

Resiliency focuses on how workloads behave during disruptive conditions such as failures, sudden changes in load, or unexpected operating stress—so they can continue operating and limit customer-visible impact. Some disruptive conditions are not “faults” in the traditional sense; elastic scale-out is a resiliency strategy for handling demand spikes even when infrastructure is healthy.

In Azure, resiliency is achieved through architectural and operational choices that tolerate faults, isolate failures, and limit their impact. Many decisions begin with failure-domain architecture: availability zones provide physical isolation within a region, zone-resilient configurations enable continued operation through zonal loss, and multi-region designs can extend operational continuity depending on routing, replication, and failover behavior.

The Reliable Web App reference architecture in the Azure Architecture Center illustrates how these principles come together through zone-resilient deployment, traffic routing, and elastic scaling paired with validation practices aligned to WAF. This reinforces a core tenet of resiliency by design: resiliency is achieved through intentional design and continuous verification, not assumed redundancy.  

Traffic management and fault isolation

Traffic management is central to resiliency behavior. Services such as Azure Load Balancer and Azure Front Door can route traffic away from unhealthy instances or regions, reducing user impact during disruption. Design guidance such as load-balancing decision trees can help teams select patterns that match their resiliency goals.

It is also important to distinguish resiliency from disaster recovery. Multi-region deployments may support high availability, fault isolation, or load distribution without necessarily meeting formal recovery objectives, depending on how failover, replication, and operational processes are implemented.

From resource checks to application-centric posture

Customers experience disruption as application outages, not as individual disk or VM failures. Resiliency must therefore be assessed and managed at the application level.

Azure’s zone resiliency experience supports this shift by grouping resources into logical application service groups, assessing risk, tracking posture over time, detecting drift, and guiding remediation with cost visibility. This turns resiliency from an assumption into an explicit, measurable posture.

Validation matters: configuration is not enough

Resiliency should be validated rather than assumed. Teams can simulate disruption through controlled drills, observe application behavior under stress, and measure continuity characteristics during expected scenarios. Strong observability is essential here: it shows how the application performs during and after drills.

Increasingly, assistive capabilities such as the Resiliency Agent (preview) in Azure Copilot help teams assess posture and guide remediation without blurring the distinction between resiliency (remaining operational through disruption) and recoverability (restoring service after disruption).  

What “enough resiliency” looks like: workloads remain functional during expected scenarios; failures are isolated, and systems degrade gracefully rather than causing customer-visible outages.

Part IV – Recoverability in practice: Restoring normal operations after disruption

Recoverability becomes relevant when disruption exceeds what resiliency mechanisms can withstand. It focuses on restoring normal operations after outages, data corruption events, or broader incidents, returning the system to a reliable state.

Recoverability strategies typically involve backup, restore, and recovery orchestration. In Azure, services such as Azure Backup and Azure Site Recovery support these scenarios, with behavior varying by service and configuration.

Recovery requirements such as Recovery Time Objective (RTO) and Recovery Point Objective (RPO) belong here. These metrics define restoration expectations after disruption, not how workloads remain operational during disruption.

Recoverability also depends on operational readiness: teams document runbooks, practice restores, verify backup integrity, and test recovery regularly, so recovery plans work under real pressure.

By separating recoverability from resiliency, teams can ensure recovery planning complements, rather than substitutes for, sound resiliency architecture.

A 30-day action plan: Turning intent into reliable outcomes

Within 30 days, translate concepts into deliberate decisions.

First, identify and classify critical workloads, confirm ownership, and define acceptable service levels and tradeoffs.

Next, assess resiliency posture against expected disruption scenarios (including zonal loss, regional failure, load spikes, and cyber disruption), validate failure-domain choices, and verify traffic management behavior. Use guardrails such as Azure Backup, Microsoft Defender for Cloud, and Microsoft Sentinel to strengthen continuity against cyberattacks.

Then, confirm recoverability paths for scenarios that exceed resiliency limits, including restoration paths and RTO/RPO targets.

Finally, align operational practices—change management, observability, governance, and continuous improvement—and validate assumptions using the Reliability guides for each Azure service.

Designing confident, reliable cloud systems

Modern cloud continuity is defined by how confidently systems perform, withstand disruption, and restore service when needed. Reliability is the outcome to design for; resiliency and recoverability are complementary strategies that make reliable operation possible.

Next step: Explore Azure Essentials for guidance and tools to build secure, resilient, cost-efficient Azure projects. To see how shared responsibility and Azure Essentials come together in practice, read Resiliency in the cloud—empowered by shared responsibility and Azure Essentials on the Microsoft Azure Blog.

For expert-led, outcome-based engagements to strengthen resiliency and operational readiness, Microsoft Unified provides end-to-end support across the Microsoft cloud. To move from guidance to execution, start your project with experts and investments through Azure Accelerate.

Azure capabilities referenced

Foundational guidance:

Get started with Microsoft Cloud Adoption Framework

Explore the Azure Well-Architected Framework

See all reliability guides in Azure services

Resiliency examples:

Read overview on Azure Resiliency

What are availability zones?

What is Azure Load Balancer?

What is Azure Front Door?

See how to use multi‑region support

Learn more about Resiliency Agent (preview) in Azure Copilot

Recoverability examples:

Protect your data with Azure Backup

Reduce risk with Azure Site Recovery

Understand redundancy, data replication, backup, and restore capabilities

Governance and validation examples:

Access Azure Monitor documentation

Read about Application Insights Experiences

Access Azure Chaos Studio documentation

What is Azure Policy?

What is Azure landing zone?

What are Azure Verified Modules?

The post Azure reliability, resiliency, and recoverability: Build continuity by design appeared first on Microsoft Azure Blog.
Quelle: Azure

Claude Sonnet 4.6 in Microsoft Foundry-Frontier Performance for Scale

Claude Sonnet 4.6 is available today in Microsoft Foundry, and it is designed for teams who want frontier performance across coding, agents, and professional work at scale.Last week, we took a major step forward with the availability of Claude Opus 4.6 in Microsoft Foundry, bringing frontier AI capable of deep reasoning, agentic workflows, and complex decision-making to enterprise developers and builders. If Opus represents the highest tier of AI performance, Sonnet 4.6 builds on that momentum by delivering nearly Opus-level intelligence at a lower price, while often being more token efficient than Claude Sonnet 4.5.

Claude Sonnet 4.6 is available today in Microsoft Foundry, and it is designed for teams who want frontier performance across coding, agents, and professional work at scale. With Sonnet 4.6, customers get access to powerful reasoning and productivity capabilities that make everyday AI a practical reality for development teams, enterprise knowledge workers, and automation scenarios.

Large Context, Adaptive Thinking, and Effort Controls

Claude Sonnet 4.6 delivers frontier intelligence at scale, built for coding, agents, and enterprise workflows.

A major highlight is its 1 million token context window (beta), matching the extended context capabilities of Claude Opus 4.6, alongside 128K maximum output. This enables teams to work across massive codebases, long financial models, multi-document analysis, and extended multi-turn workflows without fragmentation or repeated context resets.

Sonnet 4.6 also uses adaptive thinking and effort parameters, that gives Claude the freedom to think if and when it determines reasoning is required. This is an evolution from traditional extended thinking, optimizing both performance and speed. Teams can use effort parameters to better control quality-latency-cost tradeoffs.

A Developer’s Everyday Model

Claude Sonnet 4.6 is a full upgrade for software development. It is smart enough to work independently through complex codebases and handles iterative workflows without losing quality.

Enterprise software teams can expect Claude Sonnet 4.6 to deliver:

Stronger reasoning across code contextsBetter understanding of complex codebasesReliable performance across iterative development cyclesWhether you’re building features, refactoring existing modules, or debugging tricky issues, Sonnet 4.6 can follow your workflow, maintain architectural context, and adapt as you iterate.

Sonnet 4.6 is designed for back-and-forth development:

You define intentIt produces high-quality outputsYou guide refinementDeliverables stay consistent through iterationsFor teams building in Microsoft Foundry, this translates to fewer context resets, faster cycle times, and smoother development velocity.

Ref: Benchmark table published by Anthropic

Empowering High-Quality Knowledge Work

Sonnet 4.6 makes high-quality knowledge work accessible at scale, enabling teams to produce polished outputs with fewer editing cycles.

Improvements in search, analysis, and content generation make Sonnet 4.6 ideal for everyday enterprise workflows, such as:

Drafting and refining reportsSummarizing large document setsGenerating structured business documentationProducing polished presentations and narrativesConsistent quality across both single-turn tasks and extended multi-turn collaboration ensures teams spend less time refining and more time delivering.

Powerful Computer Use

Claude Sonnet 4.6 is Anthropic’s most capable computer use model yet, scoring 72.5% on OSWorld Verified. With improved precision, the model has better clicking accuracy on difficult UI elements. Claude Sonnet 4.6 enables browser automation at scale without API key dependency. It can navigate, interact, and complete tasks across any browser-based surface, including tools with no API, legacy systems, and sites you’re already logged into.

Claude Sonnet 4.6 can work across apps without explicit instruction. It can read context from one surface and act on another, checking a calendar, responding to a message, and creating an event, without the user having to orchestrate each step.

For organizations running business workflows on systems that predate modern APIs, Sonnet 4.6’s browser-based computer use is transformative. For developers, Sonnet 4.6 is a strong fit for software development workflows as a QA and testing layer. Spinning up a browser when needed, developers can delegate visual inspection and form-based validation.

Versatile Horizontal and Vertical Use Cases

Claude Sonnet 4.6 is a direct upgrade to Sonnet 4.5. Most workflows will require only minimal prompting changes.,

Search & Conversational Experiences

Sonnet 4.6 is an excellent choice for high-volume conversational products, delivering consistent quality across multi-turn exchanges while remaining cost-efficient for scale.

Agentic & Multi-Model Pipelines

Sonnet 4.6 can function as both lead agent and sub-agent in multi-model setups. Adaptive thinking, context compaction, and effort controls give developers precise orchestration tools for complex workflows.

Finance & Analytics

With stronger financial modeling intelligence and improved spreadsheet capabilities, Sonnet 4.6 is a strong fit for analysis, compliance review, and data summarization workflows where precision and iteration speed matter.

Enterprise Document & Workflow Production

Users need fewer rounds of editing to reach production-ready documents, spreadsheets, and presentations, making Claude Sonnet 4.6 a strong fit for finance, legal, and other precision-critical verticals where polish and domain accuracy matter.

Built for Scale in Microsoft Foundry

With Claude Sonnet 4.6 available in Microsoft Foundry, customers can deploy near-Opus-level intelligence within an enterprise-grade environment that supports governance, compliance, and operational tooling.

For teams building modern AI workflows, from developer assistants to enterprise automation agents, Claude Sonnet 4.6 provides a powerful, scalable foundation in Microsoft Foundry.

Try it today

And to go deeper, join us on February 23 for Model Mondays, where leaders from Anthropic will walk through both Claude Opus 4.6 and Claude Sonnet 4.6 including real-world use cases, architectural guidance, and what’s next for frontier models in enterprise deployment.
The post Claude Sonnet 4.6 in Microsoft Foundry-Frontier Performance for Scale appeared first on Microsoft Azure Blog.
Quelle: Azure

Introducing Budget Bytes: Build powerful AI apps for under $25

When developers hear “cloud” and “AI,” their first thought is often about cost. “How much will this cost me to learn? Can I build something meaningful without racking up a surprise bill?”

Budget Bytes is a new series is designed to inspire developers to build affordable, production-quality AI applications on Azure with a budget of $25 or less. Yes, you read that right, twenty-five dollars!

Budget Bytes title imageWhat is Budget Bytes?Budget Bytes is an episodic video series featuring developers building end-to-end scenarios from scratch. But here’s what makes it different:Real costs, tallied live – At the end of each episode, we show you exactly what it cost to build and run.Authentic development – Speakers show their actual process, including mistakes and debugging (because that’s real life).Practical patterns – Learn new tools, APIs, design patterns, and processes you can apply immediately.Replicable solutions – Every demo, you can look at the GitHub repository so you can deploy it yourself!This season centers around the Azure SQL Database Free Offer, demonstrating how you can leverage enterprise-grade database capabilities without the enterprise price tag.What You’ll LearnEach episode is packed with practical takeaways:

New tools and technologies – From Microsoft Foundry to Copilot Studio to Model Context ProtocolReal-world design patterns – See how experienced developers architect cost-effective solutionsHands-on deployment – Every solution can be deployed to your own Azure subscriptionContinuous learning – Each episode links to Microsoft Learn modules for deeper divesNew to Azure SQL? – Get started by learning through real use cases in each episode, and get inspired to apply them to your own ideas!

The Season LineupEpisode Date Topic Speaker(s) What You’ll Build1 January 29th, 2026 Microsoft Foundry Jasmine Greenaway AI Inventory Manager for free2 February 12th, 2026 AI-driven insurance scenarios Arvind Shyamsundar & Amar Patil Insurance AI Application3 February 26th, 2026 Agentic RAG for everyone Davide Mauri Model Context Protocol with .NET4 March 12th, 2026 Copilot Studio Integration Bob Ward AI Agents with your data using Copilot Studio for $10/month5 March 29th, 2026 Fireside Chat Wrap-Up Priya Sathy & Guests Series recap and key insightsTune inWatch the trailer here: Build Powerful AI Apps for under $25!

New episodes release regularly, each with:A full video walkthroughCompanion blog post with additional contextComplete source code on GitHubExplore the samples: Check out the Budget Bytes Samples Repository on GitHub (repositories go public with each episode release)Try the free tier: Azure SQL Database Free Offer Documentation

Ready to build on a budget? Subscribe to the Microsoft Developer YouTube channel and follow the samples repo to get notified when new episodes drop.

Budget Bytes: Real developers. Real apps. Real affordable.
The post Introducing Budget Bytes: Build powerful AI apps for under $25 appeared first on Microsoft Azure Blog.
Quelle: Azure