Google introduced two Gemini text-to-speech models, Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, on September 23.

The models let developers generate custom voices, direct dialog delivery, and replicate a speaker’s voice in Google AI Studio and through the Gemini application programming interface (API).

They are also rolling out in products including Gemini Notebook and Google Vids.

Google said the models support more than 100 languages and can be used for audiobooks, games, podcasts, and two-speaker dialog scenes.

Access starts immediately in Google AI Studio and via the API.

The release adds to Google’s speech products, which already include Cloud Text-to-Speech and are also documented under Chirp 3 HD voices.

However, Google’s documentation does not show how the new Gemini models are positioned against those existing products.

Google said voice replication requires a verbal consent recording from the voice owner that matches the reference speaker.

It added that audio generated by Gemini Audio models carries a SynthID watermark.

🔗 Source: Google

Recent Google developments