谷歌推出 Gemini 3.8 Flash / Flash-Lite 文本转语音模型，每一行台词都能精确控制

AI Digest

谷歌推出Gemini 3.8 Flash/Flash-Lite文本转语音模型，支持逐行台词控制、多语言定制及语音克隆，提升音频创作灵活性与表现力。

Google's Gemini 3.8 Flash/Flash-Lite TTS models enable line-by-line vocal control, multilingual customization, and voice cloning, enhancing audio creation flexibility and expressiveness.

Key points

• Flash TTS专注创意角色设计，支持方言转换与表演细节控制 Flash TTS focuses on creative character design with dialect conversion and performance control
• Flash-Lite TTS优化大容量场景，提供高性价比语音生成方案 Flash-Lite TTS optimizes for large-scale scenarios with cost-effective solutions
• 可扩展至2000+现成语音及自定义语音库，支持100+语言 Supports 2000+ pre-built voices and custom voice libraries across 100+ languages
• 内置语音复制保护机制，含同意验证与SynthID水印 Includes voice cloning protection with consent verification and SynthID watermarking
• 在长文本及双声场景中表现领先，适配播客与互动媒体 Outperforms competitors in long-form text and dual-voice scenarios

Takeaway: Gemini新模型通过精准控制与多语言支持，革新语音创作流程。 / Gemini's new models revolutionize voice creation through precise control and multilingual support.

Why it matters 值得关注其如何通过语音定制提升内容创作效率与沉浸感。