Android Studio Rabbit 2 opens to 40+ switchable AI coding agents

According to a Juejin article, the stable release of Android Studio Rabbit 2 went live on October 28. It is the newest stable build of Android Studio, an Android IDE open to more than 40 third-party AI agents. Rabbit 1's AI capabilities were already solid, but it had one fundamental restriction: you could only use Google's AI. Rabbit 2 removes that restriction. The IDE is no longer tied to a single AI vendor; instead it becomes a workbench for agents. You pick the agent, and the IDE supplies the tools. The article calls this BYOA, short for Bring Your Own Agent: you choose which coding agent to run inside the IDE yourself. The enabling piece is ACP, the Agent Communication Protocol, an open protocol defined by Google. Any agent that implements this protocol can connect to Android Studio. Users can freely select and switch between agents in the Agent Panel, rather than being limited to Gemini. Once connected, agents can operate the IDE's core tools directly. They see the project structure, build artifacts and run results, unlike traditional AI coding tools, such as the ChatGPT web interface, where you have to paste code over for the model to grasp the context. On this basis the article argues that an agent's ceiling is set not by the model itself but by how many IDE tools it can reach, and Rabbit 2 hands agents a full toolbox. The release also ships with official Skills guidance packages, AI change review, and LeakCanary integration. That is how far the public information goes at this point.