Latest AI news · Original source attached
Gemini 3.8 Flash TTS brings voice design tools
Google's most expressive TTS models yet: build voices from text prompts, or replicate one from 30 seconds of audio, with consent checks and SynthID.
Reported from Google (The Keyword)
Summary written from the original first-party source — the link to it is at the end of this post.
Google introduced Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, its most expressive text-to-speech models yet, describing the release as a shift from static voice presets to a dynamic creative studio. Both rolled out today across the Gemini API, AI Studio, and Google’s own products.
What’s actually new
- Voices from prompts. Flash TTS creates bespoke voices from scratch using natural language, covering more than 100 languages and dialects, alongside a library of 2,000+ production-ready voices.
- Replication with consent. Recreate a consistent voice from a 30-second audio sample, backed by verbal consent verification, SynthID watermarking, and C2PA credentials. You can also direct delivery line by line and stage a scene with two distinct speakers from a single script.
- Built for scale. Flash-Lite targets high volume dubbing and expressive voice agents at lower cost, while Flash TTS ranks first on Hume AI’s Voice Design Benchmark at 71.4.
Why it matters
Custom voices have been a walled garden in most TTS stacks. Google puts voice design directly in AI Studio and the Gemini API, with safety mechanisms built in for a technology that is easy to abuse. For creators, developers, and voice-focused tools, this lowers the barrier to producing consistent, character-driven audio at volume.
(Original source: Google (The Keyword))
AI Slop Weekly
Get this story in context, every Sunday.
The monthly rundown of what the stories above add up to — what shipped, what changed, what to actually use. 3-minute read, free forever.
Subscribe freeNo spam. Unsubscribe anytime.