can-i-run-this-llm

Can I run this LLM?

Can a NVIDIA RTX 5080 (16GB) run Gemma 3 27B?

Gemma · Gemma · estimated for a 16.0 GB machine at 4K context.

Yes, but it's tight.

Best fit: Gemma 3 27B at Q2_K · 10.5 GB · 49–73 tok/s on a NVIDIA RTX 5080 (16GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K10.5 GB49–73 tok/sTight / slow
Q3_K_M13.4 GB41–61 tok/sWon't fit
Q4_K_M16.6 GB34–52 tok/sWon't fit
Q5_K_M19.3 GB30–46 tok/sWon't fit
Q6_K22.2 GB27–41 tok/sWon't fit
Q8_028.7 GB22–32 tok/sWon't fit
FP1654.0 GB12–18 tok/sWon't fit

Ran Gemma 3 27B on a NVIDIA RTX 5080 (16GB)? Add your real numbers →