Thanks for sharing your experience with the Wonder model. We’re glad to hear it’s working well on your subject, but we totally understand the frustration with background artifacts.
You’re right that Wonder currently doesn’t support the Selection tool, but that’s a great suggestion. We’ve shared your request with the development team as part of our ongoing feedback for improving the Wonder model.
Thanks again for your patience and for sharing your suggestion earlier.
We followed up with our development team, and they’ve confirmed that, due to the way the Wonder model works, enabling both enhancement and upscaling in a single unified process, it isn’t currently possible to apply it selectively using a mask. This limitation is similar to that of the standard Upscale tool, which also doesn’t support selection-based adjustments.
That said, your request is still noted in our internal feature tracker, and we’re keeping it open in case future model updates or architectural changes make this kind of selective control more feasible.
We really appreciate you taking the time to share your use case and help guide the direction of future improvements!