Microsoft has integrated Anthropic’s Claude models into Copilot Studio, adding to the platform’s existing support for OpenAI’s large language models.

The rollout began for customers in early release environments and is scheduled to reach preview across all environments within two weeks. Production deployment is expected by the end of 2025.

The addition introduces Claude Sonnet 4 and Claude Opus 4.1 from Anthropic, which are now selectable alongside OpenAI’s models for agent orchestration, conversational workflows, and advanced reasoning.

Copilot Studio will continue to use OpenAI as the default model for new agents but now provides the option to select Anthropic models at both the orchestration layer and within prompt configuration tools.

Access controls are managed by administrators through Microsoft 365 Admin Centre, where Anthropic model use must be explicitly enabled for each tenant before becoming available.

Once activated, these models are enabled by default in the Power Platform Admin Centre, which provides further management controls.

If access to Anthropic models is revoked, agents automatically switch back to OpenAI’s GPT-4o without manual intervention.

Users can specify which model powers each workflow component, allowing selection based on specific requirements such as automation complexity or compliance logic.

Multi-model orchestration is supported, enabling agents using different primary language models to operate within the same environment.

Microsoft has indicated that feedback from users will influence future changes to model support within Copilot Studio.

Recently, Microsoft successfully tested a new data centre cooling system that surpasses existing technologies.

As reported by Catherine Bolgar on Microsoft’s official website, the innovation, known as in-chip microfluidic cooling, can dissipate heat up to three times more efficiently than traditional cold plates.