Google has introduced a new feature to its Gemini AI model, called Live Avatar, which enables users to interact with a virtual character that can talk, see, and react in real-time. This feature is initially targeted at businesses and is available through the Gemini Enterprise platform. The live avatar is designed to provide a more human-like experience, with the ability to process voice, image, and video inputs.
The Live Avatar feature uses advanced technology to generate speech, lip movements, and facial expressions in real-time, allowing for a more natural and engaging interaction. The system supports 97 languages and can switch between them automatically during a conversation. The avatar's lip movements and facial expressions are adapted to the language being spoken, providing a more realistic experience.
The live avatar can also process visual inputs from a user's camera or screen sharing, allowing it to react and respond accordingly. This feature enables businesses to provide more personalized and interactive customer service, such as virtual receptionists or customer support agents. Google has also implemented a watermarking system, called SynthID, to prevent misuse of the technology.
The Live Avatar feature is designed to be highly customizable, allowing businesses to choose from a range of pre-built avatars or create their own using a reference image and voice recording. The system aims to maintain a consistent brand identity and style, while also providing a unique and engaging experience for users.
Google has announced that the Live Avatar feature will be available to all users through the Gemini Enterprise platform, although a consumer-focused version has not been announced yet. The company is positioning the technology as a solution for businesses looking to provide more interactive and personalized customer experiences.
The Gemini 3.8 Live model, which powers the Live Avatar feature, has been ranked as one of the top-performing models for conversational speech-to-speech tasks. According to a benchmark by Artificial Analysis, the Extended Thinking model, which is a more advanced version of Gemini 3.8 Live, achieved a score of 82.6 points, while Gemini 3.8 Live ranked fifth.
The introduction of the Live Avatar feature marks a significant advancement in AI technology, enabling more natural and engaging interactions between humans and machines. As businesses continue to look for ways to improve customer experiences and provide more personalized support, features like Live Avatar are likely to become increasingly popular.
Key points
- Google's Live Avatar feature supports 97 languages and can switch between them automatically during a conversation.
- The technology is initially targeted at businesses and is available through the Gemini Enterprise platform.
- The Live Avatar feature uses advanced technology to generate speech, lip movements, and facial expressions in real-time.
