Looks like k-omega is picking up near wall very accurately and k-epsilon is picking up free stream flow accurately. Which is great and is why the SST which is a combination of both is usually a good choice. I’m not sure why your SST model isn’t mimicking the kepsilon data better though in the free stream