Those of us wishing Copilot didn’t suck just got a gift from our friends in Redmond. Starting today, Anthropic’s Claude Sonnet 4 and Claude Opus 4.1 models will be available to Copilot users in the Researcher tool and through Copilot Studio. Microsoft says this gives customers flexibility to “mix which models are used for specific tasks.”
Charles Lamanna, president of Microsoft’s business and industry Copilot team, confirmed that Copilot will still run OpenAI models by default. Users who opt in can get a “Try Claude” button in Researcher to switch between OpenAI and Claude Opus 4.1. In Copilot Studio, developers building agents can now choose from Claude Sonnet 4, Claude Opus 4.1, OpenAI models, or any model in the Azure model catalog.
Anthropic’s models remain hosted on Amazon Web Services, not Microsoft’s internal infrastructure. Microsoft accesses them via Anthropic’s API, just like you or I would. Microsoft made xAI’s Grok models available in a similar way.
This new feature follows recent internal trials in VS Code where Microsoft moved coding users to Claude Sonnet 4 by default in the editor’s automatic AI model selection. Reports suggest Claude has outperformed OpenAI in Excel and PowerPoint tasks internally, so maybe we’re going to see a broader rollout. Fingers crossed. It’s way past time for Copilot to step it up!
Author’s note: This is not a sponsored post. I am the author of this article and it expresses my own opinions. I am not, nor is my company, receiving compensation for it. This work was created with the assistance of various generative AI models.