// RUN — FITS

Can the Apple M3 Max 36GB run ProGen2-xlarge?

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

ProGen2-xlarge on the Apple M3 Max 36GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 138k 15.9-19.2 t/s ~0.5-0.6 (slow)
Q8 yes 187k 31.9-38.4 t/s ~1.1-1.3 (slow)
Q4 yes 208k 56.9-68.6 t/s ~1.9-2.3 (slow)

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

→ All models on the Apple M3 Max 36GB → ProGen2-xlarge specs → Other hardware

FAQ

Can the Apple M3 Max 36GB run ProGen2-xlarge?▶

Yes. At Q4 it generates ~56.9-68.6 tok/s and fits up to 208k context; beyond that it spills to SSD swap and slows to ~1.9-2.3 tok/s.

How much context fits?▶

At Q4, up to about 208k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.