Google introduces two new text-to-speech models in the Gemini family: Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS. The announcements describe voice generation that supports custom voice creation, including the ability to clone a voice from about 30 seconds of audio.
Google says the models are rolling out through the Gemini API and Google AI Studio starting Wednesday. NDTV reports the TTS models support 100+ languages and dialects, with Gemini 3.8 Flash aimed at “creative voice design” and Gemini 3.8 Flash-Lite positioned for high-volume audio generation. The Next Web characterizes the update as a shift from fixed voice presets toward more dynamic voice creation capabilities.
Across outlets, the focus remains on the same product release: two Gemini TTS variants, broad multilingual support, and a workflow that allows user-provided audio to inform the generated voice. Differences are mainly in emphasis, with NDTV highlighting language coverage and intended use cases, while The Next Web foregrounds the voice-cloning capability and creative framing.
