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

Can I run Phi 4 Multimodal Instruct on GeForce RTX 3090?

Yes — it runs.best quant:
Decode speed
221 tok/sest
Usable context
131K
Memory
24 GB
Fits in memory221 tok/sest · 131K ctx
Fits in memory188 tok/sest · 131K ctx
Fits in memory126 tok/sest · 66K ctx
What is quantisation? →
Phi 4 Multimodal Instruct
5.6B · full specs & benchmarks →
GeForce RTX 3090
24 GB · everything it runs →