GitHub Copilot Code Review Switches Its Default Effort Level to Balanced
Starting September 28, 2026, the "Default" effort setting for GitHub Copilot code review will point to Balanced instead of Lite for every repository and organization that hasn't explicitly picked a level. Balanced routes a pull request to a higher-reasoning model for deeper analysis of complex logic, security-sensitive code, and cross-service changes, versus Lite's faster, more targeted feedback β and it consumes more AI credits in the process. Teams that want to keep the current lighter-weight behavior must explicitly select Lite before the cutover.
Key Takeaways
- "Default" now means Balanced, not Lite, effective September 28, 2026 β a behavior change for every team that never explicitly configured an effort level.
- Balanced digs deeper into complex logic, security-sensitive code, and cross-service changes using a higher-reasoning model.
- Balanced costs more in AI credits and can use marginally more GitHub Actions minutes than Lite.
- Lite remains available and is still the better fit for low-risk changes like documentation fixes and small refactors.
- Teams must act to keep Lite β explicitly selecting it before the cutover is the only way to avoid the automatic switch.
- Org-wide defaults still cascade to repositories that haven't set their own override, so a single admin change can cover an entire organization.
What's Changing
GitHub Copilot code review ships with two effort levels, Lite and Balanced, that were made generally available earlier in August 2026. Until now, any repository or organization left on the "Default" setting was effectively running Lite. That changes on September 28, 2026: the Default value will point to Balanced instead, meaning any team that hasn't made an explicit choice will start getting a noticeably different β and more expensive β style of review without changing a single setting themselves.
Lite vs. Balanced
Lite is built for speed: fast, targeted feedback on common issues like bugs, security vulnerabilities, and style inconsistencies. Balanced trades some of that speed for depth, routing a pull request to a higher-reasoning model that spends more time on complex logic, security-sensitive code paths, and changes that span multiple services. That extra scrutiny comes at a real cost β Balanced reviews consume more AI credits and can use marginally more GitHub Actions minutes than Lite.
Why GitHub Is Making This the Default
Copilot code review's whole pitch is catching issues before a human reviewer has to. A Lite-by-default posture optimized for review speed and low overhead, which made sense while the feature was new. Shifting the default to Balanced signals GitHub now expects most teams to prefer thoroughness over speed for their baseline review, treating Lite as the opt-in choice for teams that specifically want a faster, cheaper pass β documentation fixes, small refactors, and other low-risk changes.
How to Keep Lite
Nothing about this requires action if a team is happy moving to Balanced. But any organization or repository that wants to preserve today's Lite-based behavior needs to go into its Copilot code review settings and explicitly set the effort level to Lite before September 28, 2026 β leaving it on "Default" after that date means silently switching to Balanced. Organization admins can also set an org-wide default that individual repositories inherit unless they configure their own override.