can-i-run-this-llm

Can I run this LLM?

Can a AMD Ryzen AI Max+ 395 “Strix Halo” (64GB) run DeepSeek-R1-Distill 32B?

DeepSeek · Apache 2.0 · estimated for a 64.0 GB machine at 4K context.

Yes, but it's tight.

Best fit: DeepSeek-R1-Distill 32B at Q2_K · 12.3 GB · 12–18 tok/s on a AMD Ryzen AI Max+ 395 “Strix Halo” (64GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K12.3 GB12–18 tok/sTight / slow
Q3_K_M15.9 GB10–14 tok/sTight / slow
Q4_K_M19.9 GB8–12 tok/sTight / slow
Q5_K_M23.3 GB7–10 tok/sTight / slow
Q6_K26.9 GB6–9 tok/sTight / slow
Q8_034.8 GB5–7 tok/sTight / slow
FP1665.5 GB3–4 tok/sWon't fit

Ran DeepSeek-R1-Distill 32B on a AMD Ryzen AI Max+ 395 “Strix Halo” (64GB)? Add your real numbers →