The cheapest cataloged machine that runs Phi-4 on the GPU is a NVIDIA GeForce RTX 3080 · 10GB VRAM. Here are the most affordable rigs that fit it, smallest memory first.
| machine | best fit | verdict | speed |
|---|---|---|---|
| NVIDIA GeForce RTX 3080 · 10GB VRAM | 14B Q3_K_M | ✅ Runs (tight) | 35–60 tok/s |
| AMD Radeon RX 6700 XT · 12GB VRAM | 14B Q4_K_M | ✅ Runs (tight) | 14–25 tok/s |
| AMD Radeon RX 7700 XT · 12GB VRAM | 14B Q4_K_M | ✅ Runs (tight) | 16–25 tok/s |
| NVIDIA GeForce RTX 3060 · 12GB VRAM | 14B Q4_K_M | ✅ Runs (tight) | 14–25 tok/s |
| NVIDIA GeForce RTX 4070 · 12GB VRAM | 14B Q4_K_M | ✅ Runs (tight) | 19–30 tok/s |
| AMD Radeon RX 6800 XT · 16GB VRAM | 14B Q5_K_M | ✅ Runs (tight) | 16–25 tok/s |
| AMD Radeon RX 6950 XT · 16GB VRAM | 14B Q5_K_M | ✅ Runs (tight) | 18–30 tok/s |
| AMD Radeon RX 7800 XT · 16GB VRAM | 14B Q5_K_M | ✅ Runs (tight) | 20–35 tok/s |
| NVIDIA GeForce RTX 4080 · 16GB VRAM | 14B Q5_K_M | ✅ Runs (tight) | 25–40 tok/s |
| NVIDIA GeForce RTX 5080 · 16GB VRAM | 14B Q5_K_M | ✅ Runs (tight) | 30–55 tok/s |
| Apple M3 Pro · 18GB | 14B Q3_K_M | ✅ Runs (tight) | 9–15 tok/s |
| AMD Radeon RX 7900 XT · 20GB VRAM | 14B Q8_0 | ✅ Runs (tight) | 18–30 tok/s |
Already have a machine? See Phi-4 across popular rigs → or check your exact rig →
Planning a whole workflow? Rigs for a coding agent · local chat · reasoning · local RAG setup.
Verdicts use the same GQA-aware fit math as the CLI — a NVIDIA GeForce RTX 3080 verdict here matches what runlocal tells you on the machine.