Google's new Flash TTS models let you design AI voices from scratch using text descriptions
decoder ·
Google has released two new text-to-speech models, Gemini 3.8 Flash TTS and Flash-Lite TTS, supporting over 100 languages. The standard Flash TTS allows users to design unique voices from scratch using text descriptions or clone voices from 30-second samples with consent. It is targeted at creative applications like podcasts and games, featuring stage directions and distinct two-voice dialogue capabilities. In contrast, the Flash-Lite TTS model focuses on cost-effective, large-scale speech generation for dubbing and voice agents. Both models include SynthID watermarks for detection and support nonverbal sounds such as laughter and sighs. Google is rolling out these tools via the Gemini API and Google AI Studio, with enterprise access coming soon. Pricing structures vary by model and time period, with costs increasing in 2027.
decoder