Can I run this LLM?

Can a NVIDIA RTX 4090 (24GB) run Gemma 4 31B?

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

Yes, but it's tight.

Best fit: Gemma 4 31B at Q3_K_M · 13.2 GB · 33–50 tok/s on a NVIDIA RTX 4090 (24GB).

Every quantization on this machine

Quant	Size	Speed	Fit
Q3_K_M	13.2 GB	33–50 tok/s	Tight / slow
Q4_K_M	18.6 GB	27–40 tok/s	Won't fit
Q5_K_M	21.9 GB	24–36 tok/s	Won't fit
Q6_K	25.3 GB	22–32 tok/s	Won't fit
Q8_0	32.6 GB	18–27 tok/s	Won't fit
FP16	61.4 GB	11–16 tok/s	Won't fit

Ran Gemma 4 31B on a NVIDIA RTX 4090 (24GB)? Add your real numbers →