Gemini Live on Android: Voice Control Expands With Connected Apps and a New Neural Expressive Interface

The redesign also changes how users interact with Gemini Live:

• Voice and typing now coexist in the same interface.
• Gemini Live appears directly inside the chat instead of launching a fullscreen mode.
• A floating prompt interface makes it easy to switch between text input, voice interaction, and media uploads.

The result is a more conversational assistant experience where users can start with a typed question, continue with a voice conversation, and return to text without breaking the flow.

Gemini Live now works with more Connected Apps

A key part of the update is broader support for Connected Apps, which allow Gemini to interact with other services when the user grants permission.

These integrations let the assistant move beyond answering questions and actually perform tasks across Android apps and Google services.

Examples include:

Smart home and device control

• Control smart home devices through Google Home.
• Adjust phone utilities such as alarms and device settings through Android system utilities.

Productivity and Workspace tools

• Summarize emails in Gmail.
• Create events in Google Calendar.
• Work with content in Google Docs or Drive.

Travel planning

• Retrieve flight and hotel information through Google Flights and Google Hotels for travel planning.

Media and entertainment

• Play music from YouTube Music or Spotify.
• Find videos or inspiration on YouTube.

Communication and messaging

• Make calls or send messages using default phone apps or supported messaging services.

Because these capabilities can also be used inside Gemini Live conversations, users can trigger actions with natural voice commands during an ongoing chat.

A more “agentic” Gemini assistant

The Gemini Live update is part of a larger shift Google outlined at I/O 2026: turning Gemini from a chatbot into a proactive assistant capable of completing tasks across services.

Several new components announced alongside the redesign support this strategy:

Gemini 3.5 Flash
A new generation model designed to combine strong reasoning with faster response times. It is rolling out across the Gemini app, Search, and developer tools.

Gemini Omni
A multimodal model capable of transforming text, images, or prompts into video outputs, expanding Gemini’s creative generation abilities.

Gemini Spark
A cloud-based personal agent designed to help users complete tasks continuously, positioning Gemini as an always-available digital assistant rather than a reactive chat tool.

Together, these features push Gemini toward an “agentic” model where the AI can retrieve information, generate content, and carry out actions across services on the user’s behalf.

Gemini’s rapid growth

Google says adoption of the Gemini app has surged alongside these new capabilities. By the time of I/O 2026, the app had reached around 900 million monthly active users, roughly doubling from the previous year.

That scale gives Google a large installed base to roll out the new interface, expanded integrations, and agent-style features.

Why the Gemini Live update matters

The Gemini Live redesign is less about adding a single feature and more about changing how people interact with AI on their phones.

Instead of launching a dedicated voice mode, the assistant now sits inside the main chat interface and can act across apps through Connected Apps. This combination allows users to:

• switch seamlessly between typing and voice
• control devices and services during a conversation
• retrieve information from apps like Gmail, Maps, or Flights
• trigger media playback or smart home actions

In practical terms, Google is turning Gemini into a more ambient Android assistant—one that blends conversation, app actions, and AI generation into a single continuous interface.

As Google continues expanding integrations and agent capabilities, Gemini Live may increasingly function as the central control layer for Android devices and services.