Why Google Antigravity 2.0 Split Its Most Popular AI Tools
TL;DR
Antigravity 2.0 splits Google’s agent workflow into four parts: desktop app, IDE, CLI and SDK. The app becomes the central place for scheduling, orchestration and parallel sub-agent work. The IDE is not gone, but it is no longer the default. Some updates replaced it with the new app, confusing users; projects and settings are said to remain in a separate directory.
Nauti's Take
Google’s move makes technical sense, but the rollout sounds messy. Agent orchestration, coding, terminal control and SDK-level integration are different jobs; forcing them into one product can become bloated fast.
The issue is not modularity, it is transition design: if an update makes users think their IDE disappeared, the architecture is ahead of the product communication. For teams, the real question is whether the new split clarifies ownership or just creates another tool map to maintain.
Briefingshow
The split shows where agent tools are heading: away from one all-purpose AI editor and toward a small platform made of UI, terminal, SDK and runtime logic. That can make workflows cleaner, but it also adds operational overhead. Anyone using Antigravity seriously now has to decide which tool owns which part of the job.
Video
Sources
