local intelligence

can it run

Can the NVIDIA GeForce RTX 3060 Ti run PHI4 mini 3.8B?

FULL

3.0 GB needed of 8 GB usable — headroom for context

8GB · Ampere · model file 2.5GB · mini 3.8B

~82–140 tok/s

computed roofline: 448 GB/s × 0.46–0.76 efficiency window / 2.5GB (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: ~120 tok/s · 3-4B Q4 · 12 runs · inside our computed band — localmaxxing.com

Run PHI4 mini 3.8B on instead

Raspberry Pi 5 + AI Kit (Hailo-8L) — cheapest catalog machine that runs PHI4 mini 3.8B fully ($110)

rtx-3070 — the next step up from NVIDIA GeForce RTX 3060 Ti

every model the NVIDIA GeForce RTX 3060 Ti 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 GeForce RTX 3060 Ti · fit = weights + 4k KV vs usable memory