can-i-run-this-llm

Can I run this LLM?

Can a Mac Studio M5 Ultra (96GB) run Phi-4 14B?

Phi · MIT · estimated for a 96.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Phi-4 14B at FP16 · 29.3 GB · 24–36 tok/s on a Mac Studio M5 Ultra (96GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K5.6 GB73–110 tok/sRuns great
Q3_K_M7.2 GB64–97 tok/sRuns great
Q4_K_M8.9 GB57–85 tok/sRuns great
Q5_K_M10.4 GB52–77 tok/sRuns great
Q6_K12.0 GB47–70 tok/sRuns great
Q8_015.6 GB39–58 tok/sRuns great
FP1629.3 GB24–36 tok/sRuns great

Ran Phi-4 14B on a Mac Studio M5 Ultra (96GB)? Add your real numbers →