// RUN — FITS

Can the Apple M3 Max 64GB run Codestral?

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

Codestral on the Apple M3 Max 64GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 45k 6.1-7.4 t/s ~0.2-0.2 (slow)
Q8 yes 129k 12.3-14.8 t/s ~0.3-0.4 (slow)
Q4 yes 167k 21.9-26.4 t/s ~0.5-0.7 (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 64GB → Codestral specs → Other hardware

FAQ

Can the Apple M3 Max 64GB run Codestral?▶

Yes. At Q4 it generates ~21.9-26.4 tok/s and fits up to 167k context; beyond that it spills to SSD swap and slows to ~0.5-0.7 tok/s.

How much context fits?▶

At Q4, up to about 167k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.