Google's Gemini Live Gets Camera-Triggered Images and Mid-Chat App Control
Gemini Live is now embedded directly in chat, letting you switch between voice and text mid-conversation while calling tools and generating images on the fly

- Gemini Live is now embedded in chat — no more fullscreen mode; switch between voice and text mid-conversation seamlessly.
- Live demo this Thursday at 11:30am PT on the Gemini Discord, featuring the product team walking through new capabilities.
- Mid-conversation tool connections now include Canva, OpenTable, Instacart, Spotify, YouTube, Google Workspace, and more via MCP.
- Camera-to-image generation: point your camera at something, describe what you want, and Gemini generates new images in the same voice conversation.
- Rolling out globally to Android and iOS; core features are free, advanced image models and tools require Google AI Pro ($20/mo) or Ultra ($100/mo).
- Known limitation: safety filters can be overly aggressive, sometimes blocking reasonable image generation requests.
Google is hosting a live Discord demo this Thursday to show off the next chapter of Gemini Live , and the feature set being demoed is a meaningful step forward from what most people think of when they hear "voice assistant." The session will feature Gemini product team members walking through three core capabilities: persistent conversations that blend voice and text, mid-conversation tool connections, and real-time image generation triggered by what the camera sees.
From a separate mode to a first-class citizen
The biggest structural change is that Gemini Live is no longer a separate fullscreen experience you launch and exit. Gemini Live no longer opens a fullscreen interface , it's now an inline experience where you don't have to switch between modes. Google has integrated the Gemini Live conversational experience directly into Gemini, so you can seamlessly switch from typing a quick question to diving deep into a free-flowing conversation , and back again , without missing a beat.
The mic got a hardware-level rethink too. Google re-engineered the mic so you can tap and talk through a complex idea at your own pace without getting cut off mid-thought. That's a small but important fix , the old behavior of cutting off users mid-sentence was one of the most common complaints about voice AI interfaces.
Tools, mid-conversation
The second big demo focus is tool connectivity. Previously, Gemini Live was limited to a small set of Google apps. Previously, Gemini Live only integrated with Google Calendar, Tasks, Keep, and Maps, as well as third-party equivalents from Samsung, Honor, OnePlus, Oppo, Tecno, Vivo, and Xiaomi. That list has expanded substantially.
The first-party list now includes Home, Hotels, Flights, Workspace, Image generation, Shopping, Utilities, YouTube, and YouTube Music, while there's also Spotify support. And on the third-party front, Google is expanding its list of Gemini connected apps with new MCP connections to Canva, OpenTable, and Instacart. MCP (Model Context Protocol) is an open standard that lets AI models connect to external tools through a standardized interface , think of it as a universal plugin system for AI assistants.
The practical result: you can be mid-voice-conversation and ask Gemini to book a restaurant on OpenTable, order groceries on Instacart, or pull up a Canva design , without breaking the conversational flow. Gemini now connects directly with your favorite apps to handle the heavy lifting , from comparing products to catching up on emails. While you talk, Gemini will even show you information like real-time maps, weather cards, or you can show Gemini what you're seeing to dream up new images.
Point your camera, get an image back
The third demo feature is arguably the most novel: showing Gemini what your camera sees and having it generate new images from that visual context. This is powered by Nano Banana, Google's native image generation system built into the Gemini models.
Nano Banana is the name for Gemini's native image generation capabilities. Gemini can generate and process images conversationally with text, images, video, or a combination , letting you create, edit, and iterate on visuals with unprecedented control. The Live integration means that loop is now voice-driven: describe what you want to do with what you're looking at, and Gemini generates it in the same conversation.
The conversational editing capability means you can generate an image and then tell Gemini to change specific elements , "make the sky more dramatic,