The new Gemini 2.5 audio TTS models with around 30 voices are very customizable and sound better IMO than openai (that hasn't been updated in typingmind). . more info at : https://aistudio.google.com/generate-speech

Would be very useful.