Google’s Gemini Live is learning to sound less like a robot and more like, well, you.
The company has rolled out a major update to its conversational AI, making chats on both Android and iOS smoother, livelier, and a lot more human.
“Conversations are about more than just words,” said Angela Sun, director of Product Management for the Gemini app, in a Google blog post. “They’re about the nuance in how we speak — the rise and fall of our voices, the rhythm of our sentences and the tone behind our words.”
That human-like nuance is exactly what the new Gemini Live update aims to capture. The AI can now respond with a more natural tone, pacing, and expression, even adopting different accents for a bit of personality.
The most immediate change is the new level of control you have. If Gemini is explaining a complex idea and you’re in a rush, you can simply tell it to “speed up” for a quicker summary. If you need it to slow down to really grasp the details, you can ask for that, too.
As Sun explained, this feature is all about “putting you in complete control of your learning environment.” It’s a simple change that makes learning on the go much easier.
You can now practice without the pressure
Gemini Live now doubles as a personal language coach.
You can ask it to quiz you on Korean numbers, practice casual greetings in Spanish, or rehearse phrases for real-world situations. As Google points out, this “allows you to gain real-world speaking experience in a low-risk setting, helping you build confidence before you use your skills in real life.”
For moments that matter, like preparing for a big interview or tricky chat, Gemini can now act as your sounding board. Instead of rehearsing in front of a mirror, you can run through scenarios out loud and get instant, natural responses that help you refine your delivery.
The update also adds a touch of fun. Gemini can now play characters and tell stories — from a cowboy discussing dessert ideas to Julius Caesar narrating the fall of Rome. Google says this new level of dramatization “makes learning and entertainment more immersive,” adding a playful layer to everyday AI use.
What you need to know
The Gemini 2.5 Flash Live API powers the new capabilities, and it’s worth noting that while you can customize the speed and accent of Gemini, these settings only last for the duration of the current conversation. Once you start a new session, the speech returns to the default setting.
These enhancements represent a significant step in the ongoing conversation war among AI assistants, focusing on natural, human-like speech patterns as the key to deeper user engagement.
Want more Gemini updates? Check out eWeek’s deep dive on Gemini 2.5 Flash now in general availability and what it means for developers.