can-i-run-this-llm

Can I run this LLM?

Can a Mac Studio M5 Ultra (256GB) run DeepSeek-R1-Distill 14B?

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

Yes — it runs great.

Best fit: DeepSeek-R1-Distill 14B at FP16 · 29.6 GB · 24–35 tok/s on a Mac Studio M5 Ultra (256GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K5.8 GB73–109 tok/sRuns great
Q3_K_M7.3 GB64–96 tok/sRuns great
Q4_K_M9.0 GB57–85 tok/sRuns great
Q5_K_M10.5 GB51–77 tok/sRuns great
Q6_K12.1 GB47–70 tok/sRuns great
Q8_015.7 GB39–58 tok/sRuns great
FP1629.6 GB24–35 tok/sRuns great

Ran DeepSeek-R1-Distill 14B on a Mac Studio M5 Ultra (256GB)? Add your real numbers →