// RUN — FITS

Can the Apple M3 Max 48GB run Qwen2.5-14B?

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

Qwen2.5-14B on the Apple M3 Max 48GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 75k 9.3-11.2 t/s ~0.2-0.3 (slow)
Q8 yes 165k 18.5-22.3 t/s ~0.5-0.6 (slow)
Q4 yes 204k 33-39.8 t/s ~0.8-1 (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 48GB → Qwen2.5-14B specs → Other hardware

FAQ

Can the Apple M3 Max 48GB run Qwen2.5-14B?▶

Yes. At Q4 it generates ~33-39.8 tok/s and fits up to 204k context; beyond that it spills to SSD swap and slows to ~0.8-1 tok/s.

How much context fits?▶

At Q4, up to about 204k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.