Google has launched two Gemini 3.8 Flash TTS voice models, built for direct performance scripting and high-volume audio production. The release splits speech generation between creative direction and cost-managed infrastructure.

Two Models, Two Purposes

Gemini 3.8 Flash TTS targets interactive entertainment, game development and long-form narration, where studio teams want to design a voice through written prompts. Gemini 3.8 Flash-Lite TTS focuses on automated media dubbing, customer-facing conversational agents and high-throughput translation pipelines.

Both join Google’s existing audio lineup, which previously included 3.5 Live Translate, 3.5 Transcribe, 3.8 Live and 3.8 Live Extended Thinking.

A Far Larger Voice Library

The new models replace a fixed catalogue of 30 legacy voices. Developers can now choose from more than 2,000 pre-built vocal profiles covering regional variations such as Quebec French, Scots English and Mexican Spanish, across more than 100 languages. A forthcoming voice remixing module will let audio engineers adjust timbre, pitch, pace and accent through direct text commands.

Benchmark Results

Independent evaluations place the larger model at the top of third-party rankings. On the Hume AI Voice Design Benchmark, Gemini 3.8 Flash TTS scored 71.4 overall, along with a category-leading 60.8 in accent modelling. In the Hume AI Overall Quality Index, it took first place, while Flash-Lite TTS came second, ahead of earlier baselines set by Gemini 3.1 Flash TTS.

Double-blind human trials run through Voice Arena also showed a preference for the new model in several languages, including Japanese, Brazilian Portuguese, Vietnamese, Modern Standard Arabic, Mexican Spanish and Hindi.

Built for Dialogue and Long Recordings

Multi-speaker staging and extended runs are a key focus. A single script can direct exchanges between two speakers, keeping natural turn-taking and clear vocal separation across long conversations. Audio quality and character timbre stay stable over multi-hour files, which reduces vocal drift in audiobooks and episodic podcasts.

Writers can place non-verbal markers directly in their text, using tags such as laughs, sigh or gasp. They can also add short verbal interjections like “mhm” and “yeah” to balance conversational pace.

Safeguards Against Misuse

Voice cloning requires mandatory identity checks to reduce impersonation risks. Recreating a voice needs a 30-second reference recording along with a spoken consent track from the original voice owner. Google checks that the two recordings match before processing a custom profile.

Generated files also carry an imperceptible SynthID audio watermark and cryptographic C2PA provenance metadata in the exported waveform, so downstream detection tools can identify synthetic speech.

Where It Is Available

Developers can reach both models through Google AI Studio and the standard Gemini API. They can also connect through frameworks run by Agora, LiveKit, Pipecat and Vercel. Early commercial users include Figma, HeyGen, Linguana, Wondercraft, 99.co and Ollang, which apply the models to regional media translation and customer service.

For everyday users, Flash TTS appears inside Gemini Notebook, while Google Vids uses Flash-Lite TTS. Gemini Enterprise customers will receive administrative API access in a later rollout.

What to Watch

The benchmark scores are encouraging, but real-world results will depend on how the voices perform in production. Studios and developers will want to test accent accuracy, long-form consistency and dubbing quality in their own projects. The remixing module, once released, could be an important test of how much creative control writers and audio engineers really gain.

For businesses, the split between the two models offers a practical choice: a richer, more expressive option for creative work, and a leaner one for volume. If the safeguards hold up, the release could make high-quality multilingual audio faster and cheaper to produce, while keeping clear signals about what is synthetic.

Also Read:

Qatar Urges Restraint as US and Iran Assert Control Over Hormuz

How the GCC Is Strengthening Its Position in Global Commerce