// RUN โ€” FITS

Can the Apple M4 Max 48GB run DistilProtBert?

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

DistilProtBert on the Apple M4 Max 48GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 629k 807.1-973.3 t/s ~14.8-17.8 (slow)
Q8 yes 633k 1614.3-1946.6 t/s ~29.6-35.7 (slow)
Q4 yes 634k 2882.6-3476.1 t/s ~52.8-63.7 (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 M4 Max 48GB run DistilProtBert?โ–ถ

Yes. At Q4 it generates ~2882.6-3476.1 tok/s and fits up to 634k context; beyond that it spills to SSD swap and slows to ~52.8-63.7 tok/s.

How much context fits?โ–ถ

At Q4, up to about 634k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.