Google Rolls Out Gemini Live, a Two-Way Voice Feature That Can Take on ChatGPT’s Advanced Voice Mode

Google launched several new Pixel devices at its Made By Google event on Tuesday. Alongside the hardware announcements, the tech giant also spent considerable time showcasing new artificial intelligence (AI) features for its Gemini chatbot. Among them, Gemini Live was the most significant upgrade of the evening which introduced voice capabilities to the chatbot. With this, users can have a back-and-forth conversation with the AI without needing to type or read messages. Based on its description, the feature appears to be quite similar to ChatGPT's advanced Voice Mode which was recently rolled out to some paid subscribers.

Gemini Live Rolled Out By Google

During the event, the company introduced Gemini Live as a mobile conversational experience. The feature, in essence, will allow Gemini to have natural and free-flowing conversations with voice modulations and emotions. This will make the responses generated by the AI near human-like. Google highlighted that users will get 10 different voices to choose from, each with their specific energy level, pitch, and tonality.

In a blog post, the company highlighted that Gemini Live will also offer a hands-free experience, meaning it can still listen to the user and respond verbally while in the background or when the device is locked. Google described it as being on a regular phone call.

The new mode also allows users to go back and forth over a topic with the AI and explain the context or follow-up questions better to receive better responses. Gemini Live can be interrupted mid-response if the user wants to add more information. It can also be paused if the user decides to come back to it later.

First announced at Google I/O, Gemini Live functionality is similar to OpenAI's ChatGPT Advanced Voice Mode. Interestingly, OpenAI announced the feature just one day before Google, at its Spring event. However, Google offers more options for voices and has a higher context window (one million tokens but developers have access to up to 2 million tokens), which could give it an edge. But these are early days, and we will have to wait before either of these features is widely rolled out.

Google announced that Gemini Live has started to roll out to Gemini Advanced subscribers on Android devices. At launch, the feature will only be available in English, but the company plans to expand it to more languages and bring it to iOS in the coming weeks. Notably, the Gemini Advanced subscription is part of the Google One AI Premium plan which costs Rs. 1,950 a month.