Microsoft is expanding its AI capabilities within Microsoft 365 Copilot by integrating Anthropic’s latest models, Claude Sonnet 4 and Claude Opus 4.1. The move marks a significant shift in Microsoft’s AI strategy, introducing greater model diversity beyond its long-standing partnership with OpenAI. Starting today, customers using Microsoft 365 Copilot can access Anthropic’s models in two key areas: Researcher and Microsoft Copilot Studio. While Copilot will continue to be powered primarily by OpenAI’s latest models, the addition of Claude Sonnet 4 and Opus 4.1 gives users more flexibility to choose the best AI for their specific tasks. Charles Lamanna, president of Microsoft’s business and industry Copilot team, emphasized the strategic importance of the expansion. “Copilot will continue to be powered by OpenAI’s latest models, and now our customers will have the flexibility to use Anthropic models too — starting in Researcher or when building agents in Microsoft Copilot Studio,” he said. “The addition of Claude Sonnet 4 and Claude Opus 4.1 advances our commitment to bring the best AI innovation from across the industry to Microsoft 365 Copilot, tuned for work and tailored to your business needs.” In the Researcher feature, users can now toggle between OpenAI’s deep reasoning models and Anthropic’s Claude Opus 4.1. A new “Try Claude” button appears at the top of the Microsoft 365 Copilot app, allowing users to opt in and switch seamlessly between models. This feature is initially available through the Frontier program to licensed Microsoft 365 Copilot customers who choose to participate. In Copilot Studio, developers can now build, orchestrate, and manage AI agents using Anthropic’s models. The platform also supports mixing and matching different models—including those from OpenAI and other providers in Azure’s model catalog—enabling more customized and powerful workflows. Notably, Anthropic’s models will continue to be hosted on Amazon Web Services, Microsoft’s primary cloud competitor. Microsoft accesses the models through the Anthropic API, similar to how any third-party developer would. This arrangement mirrors Microsoft’s prior agreement with xAI to host Grok 3 on Azure, suggesting future integration of Anthropic models on Microsoft’s cloud infrastructure is likely. The announcement follows a recent shift in Microsoft’s AI strategy, as the company began favoring Anthropic’s models in Visual Studio Code. GitHub Copilot users now primarily rely on Claude Sonnet 4 when using the editor’s new automatic AI model selection feature. Reports also indicate that Microsoft is evaluating Anthropic’s models for integration into Excel and PowerPoint, where early tests have shown superior performance compared to OpenAI’s offerings. “This is just the beginning — we’re committed to delivering model innovation at speed,” Lamanna said. “Stay tuned: Anthropic models will bring even more powerful experiences to Microsoft 365 Copilot.”