Anthropic brings Claude to Microsoft Foundry with Azure hosting and governance controls

Anthropic has added Azure-hosted access to Claude Opus 4.8 and Claude Haiku 4.5, with Microsoft identity, billing, networking and governance controls.

Anthropic has made Claude available in Microsoft Foundry with Azure hosting, enterprise governance controls and data residency options.

Anthropic has made Claude generally available in Microsoft Foundry, allowing organizations to run selected Claude models through their existing Microsoft Azure environment.

The launch initially covers Claude Opus 4.8 and Claude Haiku 4.5 through the Messages API. Anthropic says the models can support coding, AI agents and more complex reasoning tasks.

The Azure-hosted option works with existing Microsoft identity, networking, billing and governance systems. Customers can also choose where model processing takes place, including a US data zone for organizations with data residency requirements.

Anthropic operates the inference, the computing process that generates Claude’s responses, and acts as the data processor. Microsoft Foundry customers receive Claude usage through their Azure account rather than establishing a separate billing arrangement.

Claude is generally available now. Anthropic says more models and features will be added over time, although it has not disclosed a schedule, pricing details or a full list of supported processing locations.

Azure controls move with the Claude deployment

The new option is designed for organizations that want to use Claude while retaining the access controls and infrastructure already configured in Azure.

Administrators can use their existing Azure identity and networking arrangements rather than building a separate environment around the models. Claude use also appears on a consolidated Microsoft invoice.

Eligible organizations with a Microsoft Enterprise Agreement can apply Claude usage against an existing Azure spending commitment.

That arrangement may simplify procurement for companies, universities and public-sector organizations already using Azure, although Anthropic has not set out which customers qualify or whether additional contractual requirements apply.

The US data zone gives customers an option to keep inference processing within a defined geography. Anthropic does not identify other available data zones in the announcement.

Anthropic also makes an important distinction between Azure hosting and operation of the models. Claude runs within the customer’s Azure environment, but Anthropic continues to operate the inference service and process the associated data.

Two routes remain available through Microsoft Foundry

Organizations can access Claude in Microsoft Foundry through two deployment routes.

The first is "hosted on Azure," which provides the Azure authentication, billing, networking, governance and US data-zone options announced with the general release.

The second is "hosted on Anthropic," the route previously offered through the Microsoft Foundry preview. Anthropic says this option currently provides the full range of API features and access to models that have not yet been added to the Azure-hosted service.

Anthropic says it intends to bring the available models and features closer to parity over time. Until that happens, customers will need to choose between tighter Azure integration and broader access to Anthropic’s existing API capabilities.

Claude Opus 4.8 and Claude Haiku 4.5 launch with prompt caching and extended thinking. Prompt caching is intended to reduce repeated processing when the same background information is reused, while extended thinking gives the model more capacity to work through complex tasks before returning an answer.

NVIDIA points to enterprise AI agent use

Anthropic positions the launch around enterprise deployments of coding tools, reasoning systems and autonomous AI agents.

Justin Boitano, Vice President and General Manager of Enterprise Computing at NVIDIA, says: "At NVIDIA, we use autonomous AI agents every day to help our teams move faster and think bigger. Claude models bring strong reasoning, coding and enterprise capabilities that are valuable for complex technical work. With Claude now available in Microsoft Foundry running on NVIDIA GB300 GPUs, more organizations can run advanced, specialized AI agents with the performance, scale and security needed for production."

The announcement also includes customer statements covering software testing, nuclear safety analysis and enterprise AI services. Those examples describe individual organizations’ use of Claude and Azure rather than independently verified performance comparisons.

Anthropic has not published usage figures for the Microsoft Foundry release, service-level performance data or evidence comparing the two hosting routes.

Claude Opus 4.8 and Claude Haiku 4.5 are available in Microsoft Foundry from June 29. Organizations that need models or API functions not yet supported on Azure can continue using the Anthropic-hosted option while the two services develop toward broader parity.