Gemini 3.8 Flash TTS will clone a voice from 30 seconds — with a spoken consent clip
Google, Sep 23: Flash and Flash-Lite TTS, 2,000-plus voices, 100-plus languages. SynthID on every clip, C2PA, consent check. Rolling out in the Gemini API and AI Studio. No list prices in the post. Replication blocked in Illinois, Texas, the EEA, the UK, Switzerland and India.
Feature1 minEasy
Maya Chen1 min read
- Gemini
- DeepMind
- TTS
- Voice clone
- SynthID

Photo: GoogleSource page
Google DeepMind posted Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS on Wednesday. Leland Rechis and Alan Cowen, writing for the Gemini Audio team, say Flash is for character design and line-by-line direction; Flash-Lite is the high-volume, cheaper path for dubbing and voice agents. Both can start a voice from a 30-second sample the user owns or has rights to, after a spoken consent recording that has to match the reference speaker. Every generated clip, Google says, carries a SynthID watermark. Voice replication also gets C2PA credentials. The library is more than 2,000 production voices across more than 100 languages and dialects, including Mexican Spanish, Quebec French and Scots English.
The models are rolling out from Wednesday in the Gemini API and Google AI Studio, with Gemini Enterprise “coming soon” and Google Vids for a general audience. The post does not print list prices. Voice replication through AI Studio is not available in Illinois, Texas, the EEA, the UK, Switzerland or India. Google claims first place on Hume AI’s Voice Design Benchmark at 71.4 and a lead in accent modeling at 60.8; Flash and Flash-Lite take first and second on Hume’s Overall Quality Index. Those are vendor-cited boards. Marcello Babbili at Nota Tech Guy, filing Thursday, noted the missing prices and the lack of an independent check on whether SynthID survives a dubbing re-encode. Voice remixing — timbre, pitch, pace, accent on a library voice — is “coming soon,” with no date.


