The Cover feature i back but , it only covers human vocals. the models are still cant cover instrusments and guitar riffs. the model still cant layer vocals on a recorded instrument or reproduce same riff in a different genre.
Hello~~
Solid breakdown — and you’re right, the cover feature being “back” isn’t the same as it being fully capable across the workflows you’re describing 
The points you raised — instrument / riff covers, layering vocals on a recorded track, cross-genre riff reproduction — are all areas we’re actively developing and upgrading. They’re on the roadmap, just at different stages of progress. Specific, workflow-level feedback like yours genuinely helps shape the priority order, so thanks for taking the time ！！
While those cover-side capabilities catch up, a couple of things on Tunee worth playing with in the meantime:
Tempolor V4.6 — our most controllable music model right now for stylistic precision via prompts. Different workflow from cover, but useful for stretching style territory.
Happy Horse 1.0 — Alibaba’s latest video model, currently #1 on Artificial Analysis 1416 Elo, and at 70% off launch pricing this week.
GPT Image 2 — OpenAI’s next-gen image model, just landed on Tunee at 25% off launch pricing.
Different tools for different parts of the workflow — figured worth flagging while you’re poking around 