Compatibility check
Can I run MOSS Transcribe Preview 2B on Apple M4 Max (32-core GPU)?
Yes — it runs.best quant:
Decode speed
197 tok/sest
Usable context
33K
Memory
64 GB
Fits in memory197 tok/sest · 33K ctx
Fits in memory168 tok/sest · 33K ctx
Fits in memory113 tok/sest · 33K ctx