Compatibility check
Can I run Whisper Large v3 Turbo on Apple M5 Max (40-core GPU)?
Yes — it runs.best quant:
Decode speed
886 tok/sest
Usable context
262K
Memory
128 GB
Fits in memory886 tok/sest · 262K ctx
Fits in memory755 tok/sest · 262K ctx
Fits in memory506 tok/sest · 262K ctx