Google Unveils Customizable AI Voices for Games, Audiobooks, and More

Google has introduced two new models for speech generation: Gemini 3.8 Flash TTS and Flash-Lite TTS. The former is designed for creative projects, such as game characters and audiobooks, while the latter focuses on low-cost speech generation at scale for dubbing and voice agents.

One of the key features of these models is the ability to create new voices from text descriptions. Users can define a voice's role, accent, and vocal traits across a wide range of languages and dialects. A library of over 2,000 preset voices is also available for users who don't want to start from scratch.

Google has also announced 'Voice Remixing,' a feature that allows users to adjust the timbre, pitch, tempo, and accent of library voices. However, this feature is not yet available.