Compatibility check
Can I run gpt-oss-120b on Apple M3 Max (40-core GPU)?
Yes — it runs.best quant: Q4_K_M
Decode speed
3 tok/sest
Usable context
131K
Memory
128 GB
Q4_K_MFits in memory3 tok/sest · 131K ctx
Q5_K_MFits in memoryest3 tok/sest · 33K ctx
Q8_0Spills to system RAMest
