Stories
Gemini AI Gets Human-like Live Avatar, Responds in 97 Languages
- •Google introduces Gemini 3.8 Live with Live Avatar, enabling AI conversations with digital avatars that can see, hear, and respond in real-time.
- •Live Avatar enhances natural conversation with facial expressions and synced lip movements, making AI interactions more interactive.
- •Primarily for businesses, this feature supports customer support, virtual assistants, and interactive product demos via Gemini Enterprise.
- •AI can process audio and visual input simultaneously, understanding context and responding with voice, facial expressions, and video.
- •Supports 97 languages, adjusting lip movements and facial expressions to match language changes; businesses can create custom avatars.
Key Questions Answered
Gemini 3.8 Live with Live Avatar is primarily designed for business applications such as customer support, virtual assistants, and interactive product or service demos. The feature is currently made available through Gemini Enterprise to help organizations build more engaging, human-like interactions.
The Live Avatar can process audio and visual inputs simultaneously, allowing it to understand both user speech and on-screen visuals in real time. It supports 97 languages and can automatically adjust its lip movements and facial expressions to match the active language during a conversation.
Yes, organizations can use preset options or create custom animated avatars maintaining a specific character's style using high-quality reference images, though this capability is restricted to selected enterprise groups. Additionally, Google integrates SynthID watermarking to help users easily identify AI-generated content.
