News

Anthropic’s Claude Comes to Microsoft Foundry as Azure Adds Another Frontier Model Option

• By John K. Waters
• 06/30/2026

Anthropic’s Claude models are now generally available in Microsoft Foundry on Azure, giving enterprise customers another frontier AI option inside Microsoft’s cloud development environment.

The launch means Azure customers can build production applications and enterprise agents with Claude while using existing Azure authentication, billing, and governance tools, according to Anthropic. The company said usage will appear on customers’ Azure invoices and, for eligible customers with a Microsoft Azure Consumption Commitment, count toward that commitment.

Anthropic said the general availability release also allows inference to run on Azure infrastructure in a U.S. data region, a feature likely to matter to organizations with data residency, compliance, or procurement requirements.

The move expands a partnership announced last year by Anthropic, Microsoft, and NVIDIA, under which Anthropic agreed to scale Claude on Microsoft Azure using NVIDIA systems. Microsoft said at the time that the arrangement would broaden access to Claude and give Azure enterprise customers more model choice.

Microsoft’s documentation says developers can deploy Claude models in Microsoft Foundry, authenticate with Microsoft Entra ID or API keys, and call the Claude Messages API from Python, JavaScript, or REST. The documentation also says customers need a paid Azure subscription with a valid payment method and a billing account in a country or region where Anthropic offers the models for purchase.

For Microsoft, the availability of Claude in Foundry reinforces the company’s effort to position Foundry as a multi-model enterprise AI platform rather than a platform tied only to models from OpenAI, its longtime AI partner. Microsoft’s model catalog includes models from OpenAI, Anthropic, Meta, Mistral AI, DeepSeek, xAI, Cohere, Hugging Face, NVIDIA, Fireworks AI, and others.

For Anthropic, the move extends Claude’s reach into Microsoft-centric enterprises that may already have Azure contracts, identity systems, governance workflows, and procurement processes in place. Rather than asking those customers to adopt a separate platform, Anthropic is making Claude available through infrastructure and commercial channels many of them already use.

The launch also reflects a broader enterprise AI shift. Large customers increasingly want access to multiple frontier models, not just one provider’s model family. That can help organizations compare cost, latency, capability, risk controls, and workload fit across different systems.

The availability of Claude in Foundry may be especially relevant for companies building agents and coding tools, two areas where Anthropic has emphasized Claude’s capabilities. Microsoft’s Foundry documentation describes Claude models as useful for conversational AI, complex reasoning, code generation, and multimodal tasks, including image analysis.

The development does not mean Microsoft is moving away from OpenAI. Microsoft continues to offer OpenAI models through Azure OpenAI Service and across its AI products. But Claude’s general availability in Foundry gives enterprise developers more flexibility inside Azure, and it underscores how the largest cloud platforms are becoming marketplaces for competing AI models.

For Pure AI readers, the practical significance is less about a new model name and more about deployment. Claude is now easier for Azure customers to evaluate, procure, govern, and put into production through Microsoft’s platform. That could make model choice a more routine part of enterprise AI architecture, especially for organizations that want to avoid depending on a single frontier model provider.