What hardware runs Llama 3.2?

The cheapest cataloged machine that runs Llama 3.2 on the GPU is a AMD Radeon RX 7600 · 8GB VRAM. Here are the most affordable rigs that fit it, smallest memory first.

machine	best fit	verdict	speed
AMD Radeon RX 7600 · 8GB VRAM	3B Q8_0	✅ Runs comfortably	25–45 tok/s
Apple M1 · 8GB	1B Q8_0	✅ Runs comfortably	25–45 tok/s
Apple M2 · 8GB	1B Q8_0	✅ Runs comfortably	40–65 tok/s
Apple M3 · 8GB	1B Q8_0	✅ Runs comfortably	40–65 tok/s
NVIDIA GeForce RTX 3070 · 8GB VRAM	3B Q8_0	✅ Runs comfortably	45–75 tok/s
NVIDIA GeForce RTX 4060 · 8GB VRAM	3B Q8_0	✅ Runs comfortably	25–45 tok/s
NVIDIA GeForce RTX 3080 · 10GB VRAM	3B Q8_0	✅ Runs comfortably	75–130 tok/s
AMD Radeon RX 6700 XT · 12GB VRAM	3B Q8_0	✅ Runs comfortably	35–60 tok/s
AMD Radeon RX 7700 XT · 12GB VRAM	3B Q8_0	✅ Runs comfortably	40–70 tok/s
NVIDIA GeForce RTX 3060 · 12GB VRAM	3B Q8_0	✅ Runs comfortably	35–60 tok/s
NVIDIA GeForce RTX 4070 · 12GB VRAM	3B Q8_0	✅ Runs comfortably	50–85 tok/s
AMD Radeon RX 6800 XT · 16GB VRAM	3B Q8_0	✅ Runs comfortably	50–80 tok/s

Minimum rig by size

The smallest cataloged machine that runs each size on the GPU (default 8K context). Bigger sizes need more memory — a rig that runs the small one may not touch the large.

size	minimum rig	verdict	speed
3B	AMD Radeon RX 7600 · 8GB VRAM	✅ Runs comfortably	~25–45 tok/s (fast)
1B	AMD Radeon RX 7600 · 8GB VRAM	✅ Runs comfortably	~45–70 tok/s (very fast)

Already have a machine? See Llama 3.2 across popular rigs → or check your exact rig →

Planning a whole workflow? Rigs for a coding agent · local chat · reasoning · local RAG setup.

$npx runlocal-sh can-i-run llama-3.2 --hardware <your-rig>

Verdicts use the same GQA-aware fit math as the CLI — a AMD Radeon RX 7600 verdict here matches what runlocal tells you on the machine.