can-i-run-this-llm

Can I run this LLM?

Can a NVIDIA RTX 6000 Ada (48GB) run Kimi Linear 48B-A3B (MoE)?

Kimi · MIT · estimated for a 48.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Kimi Linear 48B-A3B (MoE) at Q5_K_M · 35.0 GB · 171–257 tok/s on a NVIDIA RTX 6000 Ada (48GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K16.0 GB238–358 tok/sRuns great
Q3_K_M21.2 GB215–323 tok/sRuns great
Q4_K_M29.8 GB186–279 tok/sRuns great
Q5_K_M35.0 GB171–257 tok/sRuns great
Q6_K40.5 GB158–238 tok/sTight / slow
Q8_052.2 GB137–205 tok/sWon't fit
FP1698.2 GB89–133 tok/sWon't fit

Ran Kimi Linear 48B-A3B (MoE) on a NVIDIA RTX 6000 Ada (48GB)? Add your real numbers →