Google Launches Gemini 3.8 Flash TTS Models With Prompt-Based Voice Design
Gemini 3.8 Flash TTS and Flash-Lite TTS introduce natural-language control for voice creation and speech delivery
Flash TTS supports custom voice generation across more than 100 languages and dialects, alongside 2,000+ production-ready voices.
Both models are rolling out through the Gemini API and Google AI Studio, with consumer access coming through Gemini Notebook and Google Vids
Sep 24, 2026 - Google has introduced Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, two text-to-speech models focused on customizable voice generation and expressive speech. The company announced the models on September 23, positioning Flash TTS for creative voice work and Flash-Lite TTS for higher-volume applications.
Gemini 3.8 Flash TTS allows users to create voices from natural-language descriptions covering characteristics such as role and accent. Google says the system supports more than 100 languages and dialects and provides access to more than 2,000 production-ready voices. It can also replicate a voice using a 30-second audio sample, subject to consent verification.
The model also provides controls for how speech is delivered. Developers can add instructions for pacing, emotion, acting cues, dialect changes, and conversational sounds. Google says Flash TTS can maintain voice characteristics across long-form audio and handle two-speaker scenes from a single script.
Flash-Lite TTS is aimed at high-volume applications such as dubbing, audio production, and voice agents. Google has not announced separate pricing for the two models in its launch post. For developers, both models are rolling out through the Gemini API and Google AI Studio. Flash TTS is also available in Gemini Notebook, while Flash-Lite TTS is being introduced in Google Vids. Enterprise API Access through Gemini Enterprise is listed as coming soon.
Google said generated audio carries its SynthID watermark. Voice replication also uses consent verification and C2PA credentials. Voice replication through AI Studio is unavailable in several regions, including India.
Sources:
https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-text-to-speech/
