September 23, 2026
Change the voiceover without rewriting the API: two Gemini 3.8 TTS models use the same interface
On September 23, Google introduced two Gemini 3.8 TTS models with a shared API schema. The company positions Flash TTS for complex dialogs and long-form speech. Until December 31, 2026, one million audio tokens cost $9 for Flash and $6 for Flash-Lite.
Google AI Studio
@googleaistudio
Introducing Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, our most expressive audio generation models to date. These models help creators, developers, and businesses produce richer, more expressive audio. Try them through the Gemini API and in AI Studio: https://aistudio.google.com/generate-speech?model=gemini-3.8-flash-tts&e=0
· 156.3K views
Previously, Gemini 3.1 Flash TTS Preview returned raw PCM. Now Gemini delivers WAV by default, and switching between Flash and Flash-Lite only requires changing the model name. During migration, style instructions and speaker labels must be moved to `speech_metadata`.
Two modes. Both models are available in Google AI Studio. Through the Gemini API, Google uses `POST /v1beta/interactions` with `response_format.type="audio"`, and the finished response can be saved as `out.wav`.
Starting January 1, 2027, Google will double the prices of both models.
