From Google · · 1m 30s
Gemini 3.8 TTS Models Launch
Google releases expressive text-to-speech models for custom voices and real-time audio creation.
In brief
Google releases expressive text-to-speech models for custom voices and real-time audio creation. Google launched Gemini 3.8 Flash TTS and Flash-Lite TTS models for creating custom voices and expressive audio experiences with built-in safety tools. Originally reported by Google.
New Gemini TTS Models
Google launched Gemini 3.8 Flash TTS and Flash-Lite TTS models for expressive audio generation across multiple platforms.
Custom Voice Creation
Create entirely new voices from scratch using natural language prompts across over 100 languages and dialects.
Voice Replication
Recreate consistent vocal profiles from a 30-second audio sample with built-in consent verification and SynthID watermarking.
Direct Performance Control
Direct line-by-line delivery with granular control over pacing, emotion, and realistic conversational sounds like <laughs> and |mhm|.
High-Volume Scale
Gemini 3.8 Flash-Lite TTS optimizes for high-volume dubbing and audio content creation with cost-efficient scale.
Global Language Support
Supports over 100 languages including Mexican Spanish, Quebec French, and Scots English for multilingual audio experiences.
Safety Features
Built-in SynthID watermarking and C2PA credentials ensure AI-generated speech is detectable and protected.
Real-World Applications
Use for audiobooks, podcasts, real-time voice agents, and interactive media with improved user experiences in Gemini Notebook and Google Vids.
Performance Metrics
Secures #1 on Hume AI’s Voice Design Benchmark and #1 for overall quality among competitors in key global languages.
Deployment Options
Available in Gemini API, Google AI Studio, Gemini Enterprise, Gemini Notebook, and Google Vids starting today.
The key point
Google launched Gemini 3.8 Flash TTS and Flash-Lite TTS models for creating custom voices and expressive audio experiences with built-in safety tools.
