local intelligence

can it run

Can the NVIDIA RTX A4000 run gemma3 12B?

FULL

9.6 GB needed of 16 GB usable — headroom for context

16GB · Ampere · model file 8.1GB · 12B

~25–42 tok/s

computed roofline: 448 GB/s × 0.46–0.76 efficiency window / 8.1GB (Q4_K_M) — bands, never points.

Run gemma3 12B on instead

Tesla P100 (16GB, used) — cheapest catalog machine that runs gemma3 12B fully ($135)

rtx-a4500 — the next step up from NVIDIA RTX A4000

every model the NVIDIA RTX A4000 can run — full list

explore the full catalog84 machines indexed · live prices · what each one can run
2026-09-17 · ← all models on the NVIDIA RTX A4000 · fit = weights + 4k KV vs usable memory