What (open) voice cloning models can be used for this use case? I have tried a few, not everyone can do songs, and they are not at the quality level than an be

What (open) voice cloning models can be used for this use case? I have tried a few, not everyone can do songs, and they are not at the quality level than an be used for song generation.