// RUN — FITS

Can the Apple M3 Max 128GB run Codestral Mamba?

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

Codestral Mamba on the Apple M3 Max 128GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 756k 18.7-22.5 t/s ~0.5-0.6 (slow)
Q8 yes 812k 37.3-45 t/s ~0.9-1.1 (slow)
Q4 yes 836k 66.7-80.4 t/s ~1.7-2 (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 128GB → Codestral Mamba specs → Other hardware

FAQ

Can the Apple M3 Max 128GB run Codestral Mamba?▶

Yes. At Q4 it generates ~66.7-80.4 tok/s and fits up to 836k context; beyond that it spills to SSD swap and slows to ~1.7-2 tok/s.

How much context fits?▶

At Q4, up to about 836k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.