Super excited GPT-5.5 is rolling out to GitHub Copilot, M365 Copilot, Copilot Studio, and Foundry today.   With deeper reasoning, stronger multistep execution, and better performance across long… | Satya Nadella | 55 comments

Super excited GPT-5.5 is rolling out to GitHub Copilot, M365 Copilot, Copilot Studio, and Foundry today.   With deeper reasoning, stronger multistep execution, and better performance across long, complex tasks, GPT-5.5 helps you go from idea to execution faster with fewer iterations to get to the right outcome.    It’s all about helping you choose the right model, or models, for the right task across your workflow. In GitHub Copilot CLI, you can move across models depending on the job.    Use faster models to scaffold and explore quickly. Use deeper reasoning models to plan, understand requirements, and make the right tradeoffs.    Then GPT-5.5 can take that plan and turn it into working code, iterating, resolving errors, using tools, and validating results along the way.    And with our Rubber Duck agent you get a multi-model reflection loop as a reviewer. GPT-5.5 can review the output of another model or vice versa, as seen in Seth Juarez's video below!   These same capabilities also show up in everyday work.    In M365 Copilot, we support multiple models including GPT-5.5 with Work IQ to reason over your work data, execute complex tasks, and generate complete documents and spreadsheets.   And in Foundry, you can build with GPT-5.5 using the security, governance, and orchestration required for enterprise applications and workflows.    Looking forward to what you build!