Google releases A2UI 1.0 protocol enabling AI agents to directly render user interfaces

Google has officially launched A2UI 1.0, the first stable version of its generative UI open protocol, which allows intelligent agents to respond not just with text, but by directly generating user interface elements. This protocol, initially announced by Google near the end of 2025, enables developers to integrate agents using SDKs for Python, TypeScript, and Dart, alongside open-source renderers for web, Android, and iOS. The A2UI framework dictates that the agent describes the desired content, while each application handles the actual UI drawing using its native components. This ensures consistent low latency and native fidelity whether the agent is embedded in a proprietary application or accessed via third-party AI assistants like Gemini or Claude. The 1.0 release specifically introduced bidirectional function calling and expanded support for accessibility and extra metadata. Open-source renderers are available for React, Angular, and Lit on the web, Flutter across mobile, web, and desktop, and native implementations using SwiftUI for iOS and Jetpack Compose for Android. Companies such as Atlassian, Alation, Bloomberg, and ByteDance have already begun utilizing the protocol for development.