According to Dynamics Beating monitoring, Google DeepMind has announced the release of Gemini 3.1 Flash TTS, which is Google's most controllable text-to-speech model to date.