can-i-run-this-llm

Can I run this LLM?

Can a NVIDIA RTX 3090 (24GB) run Qwen3.6 27B?

Qwen · Apache 2.0 · estimated for a 24.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Qwen3.6 27B at Q4_K_M · 16.3 GB · 36–54 tok/s on a NVIDIA RTX 3090 (24GB).

Every quantization on this machine

QuantSizeSpeedFit
Q3_K_M11.6 GB47–71 tok/sRuns great
Q4_K_M16.3 GB36–54 tok/sRuns great
Q5_K_M19.2 GB31–47 tok/sTight / slow
Q6_K22.2 GB27–41 tok/sWon't fit
Q8_028.6 GB22–33 tok/sWon't fit
FP1653.8 GB12–18 tok/sWon't fit

Ran Qwen3.6 27B on a NVIDIA RTX 3090 (24GB)? Add your real numbers →