3 ms·Yeah I've got the q4 gpt-oss-120b running at ~40-60 tokens per second on an M5 Pro.by datadrivenangel 7mo agoYeah I've got the q4 gpt-oss-120b running at ~40-60 tokens per second on an M5 Pro.