It is common for people to compare Copilot Chat with what they experience in Claude, especially since both solutions leverage Anthropic models in different ways within their respective platforms.

In the assessments we have conducted, Copilot appears to consume more than five times the resources or tokens that Claude uses when performing the same task. While we understand that Microsoft incorporates additional security, governance, and compliance layers that may drive higher consumption, the results from our testing suggest that the added overhead does not currently provide enough value to justify such a significant difference.

Could you share how Microsoft is addressing this area to ensure Copilot remains competitive from both a performance and efficiency perspective? It would be helpful to better understand the roadmap, optimization efforts, and any initiatives aimed at reducing resource consumption while maintaining the enterprise-grade security and governance capabilities that differentiate the platform.