local intelligence

can it run

Can the NVIDIA RTX 3090 run qwen3 4B?

FULL

3.2 GB needed of 24 GB usable — headroom for context

24GB · Ampere GA102 · model file 2.6GB · 4B

~170–270 tok/s

computed roofline: 936 GB/s × 0.46–0.76 efficiency window / 2.6GB (Q4_K_M) — bands, never points. Small models (under ~4.5B) run BELOW these bands — kernel-launch overhead keeps them from saturating bandwidth (measured against community medians, 09-06).

Community check: ~130 tok/s · 3-4B Q4 · 11 runs · just below our band (22% under the low edge) — localmaxxing.com

Run qwen3 4B on instead

Raspberry Pi 5 + AI Kit (Hailo-8L) — cheapest catalog machine that runs qwen3 4B fully ($110)

rtx-5090 — the next step up from NVIDIA RTX 3090

every model the NVIDIA RTX 3090 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 3090 · fit = weights + 4k KV vs usable memory