My impression is that each of the models (small-fx/small/medium) is used for generation of audios of varying lengths. Does anyone have a breakdown of the optimal length ranges for each model?
· Sign up or log in to comment
