Google Search Live Gets a Gemini Audio Upgrade for Smoother Replies
Talking to Google Search is starting to sound a lot less robotic.
Google is rolling out a new Gemini-powered audio upgrade to Search Live, turning voice queries into faster, more fluid, back-and-forth conversations.
In its official announcement, the tech titan says the Gemini-powered feature is now becoming available in the US on Android and iOS.
Designed for hands-free moments
Search Live is meant to work when typing isn’t practical and speed matters. Users can speak a question aloud and continue the exchange naturally, asking follow-ups without restarting a search or tapping through results.
Google says the experience is tuned for situations like fixing something mid-task, learning a new topic on the fly, or getting step-by-step guidance while your hands are busy.
The update relies on native audio responses rather than stitched-together voice output, allowing Search to adjust how it speaks based on context, slowing down for explanations or sounding more conversational during quick exchanges. Instead of feeling like a voice layer pasted onto traditional search results, Search Live supports a continuous spoken interaction that keeps pace with how people actually ask questions in the moment.
Powered by Gemini 2.5 Flash Native Audio
The new feature runs on Gemini 2.5 Flash Native Audio, Google’s updated model built specifically for live voice interactions. Rather than generating text first and converting it to speech, the model produces responses directly in audio, so replies are better paced and more consistent throughout a conversation.
According to Google, the model handles multi-turn conversations more reliably by keeping track of earlier context as users ask follow-up questions. It can also respond to more complex spoken instructions, pull in real-time information during an exchange, and incorporate that information into voice replies without interrupting the conversation.
Google extends live audio beyond everyday search
Search Live is just one of several Google surfaces now picking up built-in audio features. The updated Gemini 2.5 Flash Native Audio model is also becoming available to developers through Google AI Studio and Vertex AI for building real-time voice agents.
Google says businesses are already using the model for live customer interactions, including customer service agents that can handle multi-step conversations, follow spoken instructions, and respond naturally even in noisy environments. The native audio system also lets voice agents access real-time data during conversations without disrupting the exchange.
David Wurtz, Shopify’s VP of Product, noted that users “often forget they’re talking to AI within a minute,” crediting the new audio model for smoother merchant interactions.
UWM reported processing more than 14,000 loans through its voice system since integrating Gemini 2.5 Flash Native Audio, while Newo.ai said its receptionists can “identify the main speaker even in noisy settings” and switch languages mid-call without breaking the flow.
Aside from agents, Google is launching live speech-to-speech translation using the same audio foundation. The feature enables real-time translation while preserving tone, pacing, and pitch, and can handle continuous listening or two-way conversations across more than 70 languages.
The beta experience is launching in the Google Translate app for Android users in the US, Mexico, and India, with iOS support and additional regions planned.
Google’s latest pricing shift in India, a ₹199 AI subscription, adds new pressure to OpenAI’s dominance.
