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

Can I run Phi 4 Multimodal Instruct on GeForce RTX 5080?

Yes — it runs.best quant:
Decode speed
226 tok/sest
Usable context
66K
Memory
16 GB
Fits in memory226 tok/sest · 66K ctx
Fits in memory193 tok/sest · 66K ctx
Fits in memory129 tok/sest · 33K ctx
What is quantisation? →
Phi 4 Multimodal Instruct
5.6B · full specs & benchmarks →
GeForce RTX 5080
16 GB · everything it runs →