Anthropic's Claude Model Launches on Microsoft Azure AI Foundry in the US

en.Wedoany.com Reported - On June 29, Anthropic announced that the Claude model is now fully available on Microsoft Azure AI Foundry in the US. Enterprise users can directly deploy and invoke the Claude model within the Azure environment, leveraging existing Azure identity authentication, billing, and governance systems. The initial supported models include Claude Opus 4.8 and Claude Haiku 4.5, covering capabilities such as prompt caching and extended thinking, suitable for programming, intelligent agents, and complex reasoning scenarios.

The core change of this launch is that Claude has entered the Microsoft cloud management system already familiar to enterprises. Companies no longer need to build a separate process for model invocation, permission control, and billing management; they can complete model deployment, invocation, access control, and cost management within the Azure environment. For enterprise customers already using Azure, Claude is no longer just an external model service but an AI capability component that can be embedded into existing cloud resources, development tools, and governance processes. Software development teams can integrate Claude into code generation, code review, technical documentation processing, and intelligent agent workflows; business teams can use it for knowledge retrieval, contract summarization, customer service automation, complex document analysis, and multi-step task execution. Once the model enters the cloud platform, the chain between enterprise procurement, development, deployment, and compliance becomes shorter, accelerating application launch times.

Claude Opus 4.8 and Claude Haiku 4.5 are designed for different tasks. Opus is more suited for complex reasoning, programming, and long-task processing, while Haiku focuses on low latency, high-frequency invocation, and cost efficiency.

Prompt caching and extended thinking capabilities are two key features for enterprises when invoking Claude. Prompt caching reduces invocation costs from repeated context, making it ideal for enterprise knowledge bases, long document processing, fixed workflow tasks, and multi-turn agent invocations. Extended thinking is designed for complex problem decomposition, multi-step reasoning, code analysis, and planning tasks. For enterprise AI applications, model capability itself is only the first layer; factors such as invocation cost, response speed, permission boundaries, log auditing, and system stability truly impact deployment effectiveness. With Claude entering Azure AI Foundry, these capabilities can be combined with Microsoft Cloud's existing identity authentication, governance, and development tools, lowering the barrier for enterprises to integrate large models into production systems.

Enterprise users can also choose to run inference in the Azure environment or the Anthropic-hosted environment based on their needs. This dual-path deployment provides greater flexibility for enterprises, facilitating selection across different compliance requirements, data policies, and application scenarios.

Microsoft Azure AI Foundry in the US is becoming a key entry point for enterprises building AI applications. With Claude's addition, enterprises can compare different models within the same platform and select the appropriate capability combination based on tasks. Complex coding tasks, agent planning, and deep analysis can invoke more powerful models; high-frequency text processing, information extraction, and process automation can opt for lighter models. For industry customers in manufacturing, energy, finance, healthcare, and software development, this multi-model platform capability will influence how AI systems are built. Enterprises no longer just procure a single chat tool but select models, tools, data interfaces, and execution chains around business processes, embedding AI capabilities into R&D, operations, customer service, engineering management, and knowledge management systems.

This collaboration also indicates that the commercialization of foundational model companies is continuing to converge with enterprise cloud infrastructure. For large models to generate stable revenue, they cannot rely solely on web-based users and individual subscriptions; they must also enter enterprise development environments, cloud platforms, intelligent agent systems, and industry applications. With Claude fully available on Azure AI Foundry, Anthropic can reach more Microsoft cloud customers, and Microsoft can supplement its enterprise AI platform with a richer model selection. If both parties gradually achieve functional and model consistency in the future, Claude's deployment scope, invocation convenience, and depth of enterprise use within the Azure ecosystem will continue to expand.