Google releases Gemini text-to-speech with custom voice control
Google launched Gemini 3.8 Flash TTS and Flash-Lite TTS models for generating unlimited custom voices for audiobooks, games and podcasts.
Google launched two new text-to-speech AI tools that let creators build unlimited custom character voices from scratch. The models work across Google AI Studio, Gemini API, Gemini Enterprise, Gemini Notebook and Google Vids.
Users can control accent, emotional tone and vocal delivery. Creators can generate a high-energy DJ voice, a robotic monotone, or any character they imagine.
Once voices are created, users can direct how each line is delivered with precise control. Google says the 3.8 Flash model ranks first on Hume AI's Voice Design Benchmark for customisation and accent modelling.
The company built in safety features to prevent misuse of the voice generation technology.
- 30
- Original voices
- Unlimited/infinite
- Custom library size
- Gemini 3.8 Flash TTS and Flash-Lite TTS
- Models released
Why it mattersCustom AI voices could make audiobook and podcast production faster and cheaper. They could let game developers create richer character experiences without voice actors.
AustraliaAustralian content creators, game studios and podcast producers now have access to Google's latest voice AI tools.
✓ Claims checked against the source and corrected before publish. checked 11 h ago
