Developers and engineering teams worldwide use GitHub Copilot for high-quality, agent-powered code reviews on every pull request. We understand that any change is significant to our customers, especially when it relates to billing, so we are sharing this update early to help you plan and prepare. The sections below outline what is changing, why, and how to plan accordingly.

What’s changing

Last month, we shared how GitHub Copilot code review runs on agentic tool-calling architecture, allowing the code review agent to pull in broader repository context and produce more relevant feedback on each pull request. That agentic architecture runs on GitHub Actions using GitHub-hosted runners (Note: GitHub Copilot code review also supports self-hosted runners and GitHub-hosted larger runners which are billed at different rates than standard GitHub-hosted runners.)