// RUN โ€” FITS

Can the Apple M2 Max 96GB run ImageBind?

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

ImageBind on the Apple M2 Max 96GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 1267k 145.9-176 t/s ~3.6-4.4 (slow)
Q8 yes 1281k 291.8-351.9 t/s ~7.3-8.8 (slow)
Q4 yes 1288k 521.2-628.4 t/s ~13-15.7 (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 M2 Max 96GB run ImageBind?โ–ถ

Yes. At Q4 it generates ~521.2-628.4 tok/s and fits up to 1288k context; beyond that it spills to SSD swap and slows to ~13-15.7 tok/s.

How much context fits?โ–ถ

At Q4, up to about 1288k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.