Claude in Microsoft Foundry is now generally available

Microsoft has made Claude, Anthropic’s AI model, generally available through Azure’s AI Foundry platform. This means teams can now access Claude at production scale without building custom infrastructure, with the models running on NVIDIA’s latest Blackwell Ultra GPUs. If you’ve been experimenting with Claude through Anthropic’s API or other providers, this announcement matters because it gives you another deployment path—one that’s tightly integrated with Azure’s broader AI stack and governance tools.

Here’s what’s happening technically. When you use Claude through Azure AI Foundry, your requests run on Azure’s managed infrastructure instead of going through Anthropic’s servers directly. Microsoft is using NVIDIA GB300 Blackwell Ultra chips, which are their newest generation of GPUs designed specifically for AI workloads. This matters for performance because you get lower latency (the time between sending a request and getting a response) and higher throughput (how many requests per second the system can handle). From a practical standpoint, this translates to faster response times for your applications and the ability to handle more concurrent users without degradation. The integration with Azure AI Foundry also means you get enterprise features like role-based access control, audit logging, and cost tracking—things teams need when moving from experimentation to production.

The real value here comes down to velocity and operational simplicity. Consider a scenario where you’re building an internal documentation chatbot for your company: instead of managing Claude API keys, handling rate limits, and building custom logging, you can deploy Claude through Azure AI Foundry and immediately get integration with Azure’s monitoring, authentication, and compliance tools. You also get a faster development loop—the same platform where you’re experimenting with prompts and fine-tuning is where you deploy to production. If your team is already using Azure for compute, databases, or other services, having Claude available natively means fewer vendor juggling acts and a single bill from Microsoft. For teams building AI agents that need to handle production workloads reliably, this removes friction from the “proof of concept to production” journey that’s historically been the hard part of AI projects.