can-i-run-this-llm

Can I run this LLM?

Can a Mac Studio M5 Ultra (256GB) run Step 3.7 Flash (MoE)?

Step · Apache 2.0 · estimated for a 256.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Step 3.7 Flash (MoE) at Q5_K_M · 142.0 GB · 64–96 tok/s on a Mac Studio M5 Ultra (256GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K71.2 GB92–138 tok/sRuns great
Q3_K_M91.8 GB81–122 tok/sRuns great
Q4_K_M121.6 GB70–105 tok/sRuns great
Q5_K_M142.0 GB64–96 tok/sRuns great
Q6_K171.8 GB57–85 tok/sTight / slow
Q8_0212.0 GB49–74 tok/sWon't fit

Ran Step 3.7 Flash (MoE) on a Mac Studio M5 Ultra (256GB)? Add your real numbers →