Hi all, I'm looking for the best model for a hobby project and trying to make sense of the various data I came across. I know benchmarks do not often translate to the real world, especially to your particular use case (whatever it may be). But this is truly baffling: AA says Qwen 3.8 27B is better b
