// RUN ON APPLE

Run gpt-sw3-40b on Apple hardware

17 options ranked by estimated speed at Q4 / 8k context.

HardwareEst tok/sFits 8k?Max ctx~Price
Apple M2 Ultra 192GB โ†’ ~29.3 yes 454k ~$6,600
Apple M2 Ultra 64GB โ†’ ~29.3 yes 103k ~$4,000
Apple M1 Ultra 128GB โ†’ ~29.3 yes 279k ~$4,800
Apple M1 Ultra 64GB โ†’ ~29.3 yes 103k ~$3,700
Apple M4 Max 64GB โ†’ ~20 yes 103k ~$3,900
Apple M4 Max 36GB โ†’ ~20 yes 26k ~$3,000
Apple M4 Max 128GB โ†’ ~20 yes 279k ~$4,900
Apple M4 Max 48GB โ†’ ~20 yes 59k ~$3,300
Apple M1 Max 32GB โ†’ ~14.6 yes 15k ~$2,000
Apple M3 Max 48GB โ†’ ~14.6 yes 59k ~$3,200
Apple M3 Max 64GB โ†’ ~14.6 yes 103k ~$3,700
Apple M3 Max 128GB โ†’ ~14.6 yes 279k ~$4,700
Apple M1 Max 64GB โ†’ ~14.6 yes 103k ~$2,600
Apple M2 Max 32GB โ†’ ~14.6 yes 15k ~$2,200
Apple M2 Max 64GB โ†’ ~14.6 yes 103k ~$2,800
Apple M2 Max 96GB โ†’ ~14.6 yes 191k ~$3,300
Apple M3 Max 36GB โ†’ ~11 yes 26k ~$2,900

Estimate at Q4 (memory-bound). Click any option for its full per-hardware estimate and measured numbers.

FAQ

What Apple hardware runs gpt-sw3-40b?โ–ถ

On GenAIList, 17 Apple options can run gpt-sw3-40b at Q4 โ€” the fastest is the Apple M2 Ultra 192GB at ~29.3 tok/s.

How fast is gpt-sw3-40b on Apple hardware?โ–ถ

Estimated generation speed ranges from ~11 to ~29.3 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.