Gemini 3.8 text-to-speech says hello
Google is moving Gemini speech generation from preset voices to directed, prompt-built performances.
Gemini 3.8 Flash TTS can create custom voices from natural-language prompts, replicate authorized voices from 30-second samples, and control delivery line by line. Flash-Lite is framed as the cheaper, high-volume option for dubbing, content creation, and voice agents. Google says generated audio includes SynthID watermarking, with consent checks for voice replication. The models are rolling out in Gemini API, Google AI Studio, Gemini Notebook, and Google Vids, with enterprise API access coming later. Google DeepMind's note
Gemini 3.8 Flash TTS can create custom voices from natural-language prompts, replicate authorized voices from 30-second samples, and control delivery line by line. Flash-Lite is framed as the cheaper, high-volume option for dubbing, content creation, and voice agents. Google says generated audio includes SynthID watermarking, with consent checks for voice replication. The models are rolling out in Gemini API, Google AI Studio, Gemini Notebook, and Google Vids, with enterprise API access coming later. Google DeepMind's note
score 7