The way AI-generated content is delivered is evolving from dry Markdown characters into dynamic, living layouts.
At I/O 2026, Google announced a major interface overhaul for Gemini, rolling out to its 900 million monthly active users. Dubbed Neural Expressive, the new design language brings smooth animations, haptic feedback, and native integration of Gemini Live (two-way voice calling). It also introduces Tailored Responses — replies that are rendered in real time with custom layouts tailored to the query.
The update is all about killing the dreaded wall of text. Instead of spitting out dense paragraphs, Gemini now generates responses that combine high-res images, draggable interactive timelines, short videos with narration, and live data charts — all laid out dynamically based on what you asked.
On the voice front, the new UI puts Gemini Live's seamless two-way conversation experience front and center, letting users switch between typing and freeform voice chat at any time.
To make conversations feel more natural, the pause detection logic on the microphone side has been redesigned. It now allows longer pauses while users organize complex thoughts, preventing the AI from cutting in prematurely. Gemini Live will also gradually add dialect options in the near future.
The Neural Expressive design language begins rolling out today globally on the Web, Android, and iOS.