Is there likely to be a way to stream this audio in the future? As in, here's an incoming stream of text, generate the audio on-the-fly instead of all at once.