tech π€ Google lets users create custom TTS voices, unlike OpenAI
Google released Gemini 3.8 Flash TTS and Flash-Lite TTS through the Gemini API and Google AI Studio. This allows developers to create custom voices by describing them or using short recordings from the same speaker. Voice replication requires two clean samples and a consent recording, which Google verifies before issuing a voice ID. Alternatively, voice design lets users create a persona from a natural-language description across 130 supported languages. Gemini 3.8 processes input text as a verbatim transcript, a key change from previous versions. π€