Exploring GPT‑5.1 in Microsoft Copilot Studio: What You Need to Know

Microsoft has introduced GPT‑5.1 to Copilot Studio, aligning with OpenAI’s latest release. Currently, this model is available as an experimental feature for U.S. customers using early release cycle Power Platform environments. If you’re already building agents in Copilot Studio and have access to the right environment, you can start testing GPT‑5.1 today.

So, what makes GPT‑5.1 noteworthy?

One of its key enhancements is improved adaptability in “thinking time” across both chat and reasoning scenarios. This flexibility matters because agent experiences can range from quick conversational responses to complex, multi-step workflows. By adjusting how long the model “thinks,” GPT‑5.1 can deliver higher-quality results across diverse use cases.

Because GPT‑5.1 is still experimental, Microsoft recommends using it in non-production environments while internal evaluations continue. This early access offers a unique opportunity to explore cutting-edge capabilities, compare performance with existing models, and prepare for future deployments—all while Microsoft ensures quality and readiness for broader availability.