Hello! We are currently working on a new product for which we will need to generate hundreds of short audio stories. We are generally quite happy with the current quality of the Gemini 2.5 Pro Preview TTS Model Multispeaker, but it is very unstable.
Speaker voices keep switching and voices change from generation to generation. We plan to release v01 at the beginning of next year and we need a stable TTS model for this. Is there an expected timeline for the stable release of Gemini 2.5 Pro TTS?