// RUN โ€” FITS

Can the Apple M3 Max 48GB run Playground v2.5?

Yes โ€” here's how fast and up to what context, with real measured numbers.

Playground v2.5 on the Apple M3 Max 48GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 529k 38.9-46.9 t/s ~1-1.2 (slow)
Q8 yes 583k 77.7-93.7 t/s ~1.9-2.3 (slow)
Q4 yes 606k 138.8-167.3 t/s ~3.5-4.2 (slow)

Estimate (memory-bound). Beyond "max context" the KV cache spills to SSD swap and speed drops to the "beyond" figure.

FAQ

Can the Apple M3 Max 48GB run Playground v2.5?โ–ถ

Yes. At Q4 it generates ~138.8-167.3 tok/s and fits up to 606k context; beyond that it spills to SSD swap and slows to ~3.5-4.2 tok/s.

How much context fits?โ–ถ

At Q4, up to about 606k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.