Microsoft has begun implementing the AI models Claude Sonnet 4 and Claude Opus 4.1 from Anthropic in Microsoft 365 Copilot. Now, business users of Copilot can choose between Anthropic and OpenAI models for conducting complex research, creating tools, and developing corporate agents. Claude Opus 4.1 is designed for tasks requiring deep analysis, complex coding, and architecture planning, while Claude Sonnet 4 is for routine development, processing large volumes of data, and content generation.
The new models are already available to organizations with a Copilot license that have joined the Frontier Program and agreed to participate in testing. The Microsoft 365 Copilot app now features a “Try Claude” button, allowing users to switch between OpenAI and Anthropic models in “Researcher” and Copilot Studio. Users can select and combine models for different tasks, tailoring the tool to business needs.
Claude Sonnet 4 and Opus 4.1 are also available for creating and managing corporate agents in Copilot Studio. This opens up the possibility of automating workflows and solving specialized tasks using different AI models. Access to Anthropic models is provided via API, and they are hosted on Amazon Web Services servers.
The decision to add Anthropic to Microsoft 365 Copilot was made after Anthropic models showed better results in certain scenarios compared to OpenAI in tests. Microsoft continues to expand AI options for corporate users to provide more flexibility and opportunities for personalizing work tools.
