Models /Phi 4 Multimodal Instruct /Can I run it?

Compatibility check

Can I run Phi 4 Multimodal Instruct on Apple M3 Ultra (80-core GPU)?

Yes — it runs.best quant:

Decode speed

169 tok/sest

Usable context

131K

Memory

512 GB

Fits in memory169 tok/sest · 131K ctx

Fits in memory144 tok/sest · 131K ctx

Fits in memory96 tok/sest · 131K ctx

What is quantisation? →

Phi 4 Multimodal Instruct

5.6B · full specs & benchmarks →

Apple M3 Ultra (80-core GPU)

512 GB · everything it runs →