Microsoft named a Leader in the 2026 Gartner® Magic Quadrant for Container Management

I’m excited to announce that Microsoft has been recognised as a Leader in the 2026 Gartner® Magic Quadrant for Container Management, ranking highest for Completeness of Vision. We believe this accolade showcases our capability to assist customers in modernising their applications and adopting AI workloads without complicating their operations.

This recognition comes at a crucial time when container platforms are being challenged to accommodate a diverse array of workloads, operational models, and deployment environments that organisations might not have foreseen.

Over a decade ago, when we started working on Kubernetes, our initial goal was straightforward: to make distributed systems accessible, thereby simplifying the creation of reliable services. We aimed to describe workloads based on their needs rather than their placement, allowing the scheduler the freedom to make informed decisions. This approach turned out to be more significant than we realised, as it meant the system remained flexible regarding the workloads.

The rise of AI has significantly altered the landscape of container management. While Kubernetes has shown itself capable of handling AI workloads well, the key change is that applications and AI now require access closer to the data and users, often confined within specific regulatory boundaries. Today, organisations need more than just container orchestration; they require a platform that ensures a consistent operating model across cloud, edge, and hybrid deployments. This need to adapt quickly without the necessity of rebuilding applications is core to Microsoft’s container offerings, including Azure Kubernetes Service (AKS), Azure Container Apps, Azure Arc, and Azure Kubernetes Fleet Manager.

Utilise AI on Your Existing Platform

Through our customer implementations, we’ve noticed two primary architectural models emerging. In the first model, a platform team manages a smooth operational layer, overseeing GPU scheduling, model life cycles, and the necessary compliance boundaries. As demand increases and stabilises, organisations often want their AI infrastructure to function like any other platform capability: application teams utilise it while platform teams maintain control over its governance and usage. Tools such as the AI toolchain operator on AKS help automate model deployment and GPU allocation, while AKS’s CNCF AI Conformance certification assures customers that the surrounding ecosystem remains compatible as it evolves.

In the second model, an application or agent triggers inference as needed, executes generated code, and releases resources once the task is completed. This structure emphasises elasticity and isolation, requiring resources to be deployed swiftly and removed when no longer needed, while safely managing unpredictable workloads. Azure Container Apps is tailored for this operational model, providing serverless GPUs for immediate inference and hardware-isolated environments for agent hosting that maintains state throughout interactions.

Almost every enterprise we engage with requires both models, and I believe that the fascinating engineering challenge lies in creating a seamless transition between them: the same image, identity, and network controls, regardless of which model a team adopts. Platform teams appreciate the governance from the first model, while application teams and agent frameworks prefer the flexibility of the second, often wanting to avoid the complexities of Kubernetes.

Maintain a Single Operating Model as Your Estate Expands

As inference migrates closer to data sources, your environment may no longer be centralised. Clusters can emerge across various regions, data centres, locations, or in circumstances where connectivity is unreliable or restricted due to sovereignty regulations that require workloads and data to remain within specific jurisdictions. These scenarios usually lead to coordination issues rather than just cluster failures: configuration drifts, uneven upgrades, and inconsistent policy application across environments. Often, hybrid strategies falter when teams view coordination challenges as isolated cluster concerns instead of platform-wide issues.

Understanding that AI must extend from cloud to edge, we’ve developed AKS Everywhere to deliver a consistent, Azure-driven, and secure Kubernetes platform from the cloud to edge. Furthermore, through Azure Arc for Kubernetes, we extend a unified identity, policy, and observability model across CNCF-compliant Kubernetes environments, even those in other clouds. With numerous clusters, managing them effectively can lead to what we call cluster sprawl. Azure Kubernetes Fleet Manager helps tackle the coordination challenges that come with growth, aiding organisations in managing upgrades, workload placement, and consistent policy across fleets.

Ensuring everything functions smoothly relies on AKS staying close to upstream Kubernetes, which we have intentionally maintained. There’s no proprietary version, and open-source principles are central to our strategy. Notably, Microsoft has become the second-largest contributor to CNCF projects overall and the leading contributor among cloud providers for the past three years. This contribution stabilises the API you work with, regardless of where the workload may reside, and ensures that the surrounding ecosystem remains consistent both within and outside Azure.

Maintain Steady Operations as Your Estate Grows

Cluster numbers frequently increase faster than operational teams can handle, and many organisations experience this challenge before having a clear plan in place.

One solution lies in implementing better defaults. AKS Automatic applies operational approaches based on Microsoft’s extensive experience managing Kubernetes at scale, all while maintaining the flexibility of the Kubernetes API.

However, a broader transformation is shifting towards agent-centric operations. I believe this aspect will see the most change in the coming years. The Azure SRE Agent and the AKS MCP Server assist operators in moving from alert to diagnosis to remediation, utilising the same permissions and controls they currently have. The intention here isn’t to replace operators but to alleviate the burden of routine investigations that often consume excessive operational time.

Building a platform that can adapt to diverse new requirements necessitates more insight from those managing it due to its increased surface area. Our aim is to continuously simplify this complexity in the platform, and there’s still much work ahead.

Customer Progress

The following examples illustrate how customers leverage Azure’s container offerings across AI, mission-critical applications, and hybrid environments.

• Wayve employs AKS to train its autonomous driving models on vast amounts of video and sensor data, aggregating thousands of GPUs into a flexible training environment.
• AT&T developed Ask AT&T using AKS as the orchestration backbone for its containerised agents, ensuring that each agent passes through legal, security, and finance checks before production.
• Replit creates applications based on simple language descriptions and deploys them as Container Apps in the customer’s Azure environment, integrating their network and compliance measures right from the initial deployment. About 75% of their enterprise users aren’t professional coders.
• SimCorp transitioned its investment management platform, used by many leading asset managers, from virtual machines to AKS to ensure consistent identity, logging, and security policies across all clients within every jurisdiction, with a focus on auditability.
• Emirates Global Aluminium operates about two-thirds of its estate in Azure and the remainder on its own sites, processing image and video analytics directly at the plant’s floor, while allowing applications to shift seamlessly between both environments without major redesigns.

A training cluster with thousands of GPUs looks quite different from a regulated multi-tenant SaaS platform or analytics operations next to a smelter. What unites them is that each eventually required something that their initial design could not accommodate.

Azure Kubernetes Service provides the control and cost efficiency we need. We can scale GPU resources according to demand and test new models without affecting production.

Brian Sutliffe, Vice President of Engineering, CallRevu

Being acknowledged as a Leader in this Magic Quadrant is a significant achievement for us, setting clear expectations for our future endeavours. Ultimately, the team managing a workload must determine its best location, and this decision can differ across an estate. The platform’s role is to allow that choice to evolve without necessitating the redesign of the application or the adoption of a new operating model. I’d like to extend my heartfelt gratitude to all involved in Azure Cloud Native, as this recognition truly reflects the hard work of many individuals across Microsoft and Azure.

If you’d like to see the full details, you can access a complimentary copy of the 2026 Gartner® Magic Quadrant for Container Management here.

Gartner ® Magic Quadrant for Container Management, Dennis Smith, Tony Iams, Wataru Katsurashima, Lucas Albuquerque, 2 September 2026

Gartner does not endorse any company, vendor, product, or service depicted in its publications and does not advise technology users to choose only those vendors with the highest ratings or other distinctions. Gartner publications represent the views of Gartner’s business and technology insights organisation and should not be regarded as definitive statements. Gartner disclaims all warranties, expressed or implied, regarding this publication, including any warranties of merchantability or fitness for a specific purpose.

Gartner and Magic Quadrant are registered trademarks of Gartner, Inc., and/or its affiliates.

This graphic was published by Gartner, Inc. as part of a broader research document and should be considered within the context of the entire document. The Gartner document is available upon request here.