GitHub Copilot introduces Lite and Balanced code review effort levels

GitHub has released the general availability of "effort levels" for its Copilot code review feature. This update allows developers to choose the depth of AI-driven analysis based on the complexity and risk of a pull request.

There are two available levels:

- Lite: The default setting, intended for routine, straightforward changes such as documentation updates or small bug fixes. - Balanced: Designed for large-scale, complex, or security-sensitive changes. It utilizes a model with higher reasoning capabilities to perform a deeper analysis across multiple services or complex logic.

Choosing the Balanced level consumes more GitHub Actions execution time and AI credits compared to Lite. Settings can be managed at the organizational or repository level, with repository-specific configurations overriding organizational defaults. Individual users may also select the depth for specific reviews. The level used for each review is recorded in the pull request timeline to provide transparency regarding the depth of the automated inspection.