█

AI/TLDR

Google · 2026-11-11 · major

A2UI 1.0 — Google's generative UI protocol reaches its first stable version

In short

A2UI 1.0 is the first stable version of Google's open protocol for generative UI. An agent describes a screen, and each app draws it with its own native components. 1.0 adds two-way function calls, accessibility and new agent SDKs.

Google · 2026-11-11 · 4 sources checked

Google Developers Blog header image for the A2UI 1.0 announcement

Agents send a description of a screen; React, Flutter, SwiftUI and Compose apps render it with their own components.

Key specs

GitHub stars17,159

Quick facts

MakerGoogle, open-source project
Version1.0 (first stable)
LicenseApache-2.0
RenderersReact, Angular, Lit, Flutter, SwiftUI, Jetpack Compose (0.9.1)
Agent SDKsPython, TypeScript, Dart
What's newTwo-way function calls, accessibility, extensions

What is it?

A2UI 1.0 freezes Google's open generative UI protocol as a stable spec. The release adds bidirectional function calls, so a date picker can check availability in real time, plus accessibility support and extensions for extra metadata.

How does it work?

In A2UI, the agent never ships code: it streams a declarative description of the interface, and the client draws it from its own trusted component set. Data and child components are bound on the server, so sensitive values never reach the model. Models can compose the UI through A2UI Express or through decision models that pick components instead of writing text.

Why does it matter?

A stable spec means one agent can drive native screens in web, iOS, Android and Flutter apps without per-platform UI code. Google says Atlassian, Alation, Bloomberg and ByteDance already build with A2UI, and Gemini Enterprise is moving to 1.0.

Who is it for?

developers building agent apps with rich UI

Frequently asked questions

How do I upgrade from A2UI 0.9 to A2UI 1.0?
Google publishes an evolution guide in the A2UI repository that lists every change from version 0.9 to version 1.0. Some platforms still run 0.9 or 0.9.1 today: Gemini renders A2UI 0.9.1, the Jetpack Compose renderer is at v0.9.1 with 1.0 coming soon, and Gemini Enterprise is being upgraded to A2UI 1.0.
Which platforms can render A2UI 1.0 screens?
A2UI 1.0 ships open-source renderers for React, Angular and Lit on the web, Flutter for mobile, web and desktop, SwiftUI for iOS, and Jetpack Compose for Android. ByteDance also built a Lynx renderer. Each renderer draws the agent's screen description with that platform's own native components.
Does A2UI send my private data to the language model?
No. A2UI binds the interface to data and child components on the server, so Google says sensitive values never reach the model. The agent only describes which components to show, and the app fills them in with real data as the UI streams in while the agent works.
What are A2UI Macros?
A2UI Macros are a Python-only feature of the A2UI agent SDK that expands components programmatically and handles data binding, so an agent does not have to spell out every repeated component by hand. The TypeScript and Dart agent SDKs that arrive with A2UI 1.0 cover skill generation, message translation and error correction.

Try it

https://a2ui.org/quickstart/

Sources · 4 outlets

Tags

  • a2ui
  • generative-ui
  • agents
  • protocol
  • google
  • react
  • flutter
  • swiftui
  • jetpack-compose
  • open-source
  • apache-2-0

← All releases · Learn AI