AI

Google expands Antigravity and Gemini API to power agentic app development

Tuesday, May 19, 2026Read Original

Details

  • At Google I/O 2026, Google introduced Gemini 3.5 Flash, claiming it outperforms Gemini 3.1 Pro on most benchmarks while running roughly four times faster than other frontier models, to power high-speed agentic workflows.
  • The company launched the Antigravity 2.0 desktop app, Antigravity CLI, Antigravity SDK and Antigravity integration in the Gemini Enterprise Agent Platform, creating a unified agent-first development ecosystem that spans desktop, terminal, SDK and Google Cloud.
  • New Managed Agents in the Gemini API let developers spin up hosted agents that reason, use tools and execute code in persistent, isolated Linux environments, powered by the Antigravity agent harness and Gemini 3.5 Flash, with resumable state across multi-turn sessions.
  • Google AI Studio gained a pre-registerable mobile app, deeper integrations (including Google Workspace APIs and one-click export to Antigravity) and native Android support, allowing prompt-based generation of Kotlin apps with Google Play Console test-track publishing.
  • Google introduced a $100/month Google AI Ultra subscription with 5x higher Antigravity usage limits than Google AI Pro and announced the Build with Gemini XPRIZE Hackathon, a global competition with a $2 million prize pool culminating at the Moonshot Gathering in Los Angeles in September 2026.

Impact

By pairing a faster multimodal model with a production-grade agent harness and managed execution environments, Google is moving developers from ad hoc prompt-based coding toward persistent, workflow-centric agents. Over the next 12–24 months, this stack is likely to intensify competition around agent platforms, shift some dev tooling spend toward hosted agent runtimes, and accelerate adoption of AI-driven software development on Android, Workspace and Google Cloud.

Rift Dispatchpractical systems & stories, weekly