// RUN — FITS

Can the Apple M3 Max 36GB run Transformer-XL (257M)?

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

Transformer-XL (257M) on the Apple M3 Max 36GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 464k 396.9-478.6 t/s ~13.2-16 (slow)
Q8 yes 468k 793.8-957.2 t/s ~26.5-31.9 (slow)
Q4 yes 469k 1417.5-1709.3 t/s ~47.2-57 (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 36GB → Transformer-XL (257M) specs → Other hardware

FAQ

Can the Apple M3 Max 36GB run Transformer-XL (257M)?▶

Yes. At Q4 it generates ~1417.5-1709.3 tok/s and fits up to 469k context; beyond that it spills to SSD swap and slows to ~47.2-57 tok/s.

How much context fits?▶

At Q4, up to about 469k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.