What hardware runs Command R?

The smallest-memory cataloged machine that runs Command R on the GPU is a NVIDIA GeForce RTX 5090 · 32GB VRAM. Here are the rigs that fit it, smallest memory first.

machinebest fitverdictspeed
NVIDIA GeForce RTX 5090 · 32GB VRAM 35B Q3_K_M ✅ Runs (tight) 40–70 tok/s
AMD Radeon Pro W7900 · 48GB VRAM 35B Q5_K_M ✅ Runs comfortably 14–25 tok/s
Apple M3 Max · 48GB 35B Q4_K_M ✅ Runs (tight) 8–14 tok/s
Apple M4 Max · 48GB 35B Q4_K_M ✅ Runs (tight) 10–17 tok/s
Apple M4 Pro · 48GB 35B Q4_K_M ✅ Runs (tight) 6–11 tok/s
Apple M1 Max · 64GB 35B Q5_K_M ✅ Runs comfortably 7–12 tok/s
Apple M1 Ultra · 64GB 35B Q5_K_M ✅ Runs comfortably 13–20 tok/s
Apple M2 Max · 64GB 35B Q5_K_M ✅ Runs comfortably 7–12 tok/s
Apple M2 Ultra · 64GB 35B Q5_K_M ✅ Runs comfortably 13–20 tok/s
Apple M2 Max · 96GB 35B Q8_0 ✅ Runs comfortably 5–8 tok/s
Apple M2 Ultra · 128GB 35B Q8_0 ✅ Runs comfortably 9–15 tok/s
Apple M3 Max · 128GB 35B Q8_0 ✅ Runs comfortably 5–8 tok/s

Already have a machine? See Command R 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 command-r --hardware <your-rig>

Verdicts use the same GQA-aware fit math as the CLI — a NVIDIA GeForce RTX 5090 verdict here matches what runlocal tells you on the machine.