Muse Spark 1.3 supports Text, Image, Video, File, Audio inputs. We've only started to see models from China include image and video inputs recently.
I do not trust any provider, US or Chinese when they say they will not train on my data. I still use these services, but I am under no illusion that any of these people are trustworthy bunch.
While I agree with you for the Muse Spark as hosted by Meta, if it will be available in open weights form for self hosting, then there are good chances that it can become quite useful.