Google AI Studio adds multiplayer coding
Google’s AI Studio now supports multiplayer coding, real services integration, persistent builds, and pro UI components — features meant to accelerate rapid prototyping and portfolio app development. Alongside new interactive tools like AITutor (an interactive terminal for learning AI coding), the updates lower friction for building deployable projects. ( )
Google’s new Antigravity coding agent will proactively detect when an app needs backend services and—once the developer approves—provision Cloud Firestore for data storage and Firebase Authentication for Google sign‑in. (blog.google) AI Studio now includes a server‑side Node.js runtime that can install npm packages, manage secrets via a built‑in Secrets Manager, and run server logic that keeps API keys off the client; the platform documentation documents server‑side runtimes and secrets handling. (ai.google.dev) Build mode explicitly supports real‑time multiplayer experiences such as collaborative workspaces and games, and introduces a “pick up where you left off” persistent runtime that preserves app state across sessions and devices. (blog.google) The Antigravity agent will automatically add production UI libraries—Google’s blog cites examples like Framer Motion and shadcn—to create smoother animations and professional icons, and Google maintains a stitch‑skills repository with a dedicated shadcn‑ui skill for integrating those components. (blog.google) AITutor referenced alongside the release is an open‑source, terminal‑based interactive guide for AI coding with a public GitHub repository and an npm CLI package, and the project has been highlighted in community curation posts. (github.com) Google’s docs and codelabs show generated apps can be exported and deployed (for example, to Cloud Run and GitHub), and the update arrives as Gemini 3.1 Pro is being surfaced to developers in AI Studio; multiple outlets report the full‑stack vibe coding rollout on March 18–19, 2026. (codelabs.developers.google.com)
