My M5 MacBook Air arrived today, and I collected together all the computers I have access to and ran some benchmarks with an Xcode project of ours that's long enough to bring out thermal constraints. Ran two clean builds back-to-back, and these are the results from the second build (which were universally faster than the first).
It's weird that the Mac mini couldn't beat the M4 Air, but it *is* on Sequoia and everything else is Tahoe. Everything is running Xcode 26.2.


