Can I run Mistral Small 3?

24B that punches at 70B level with much faster generation. 32K context.

chatgeneralreasoningagents · sizes: 24B · Apache-2.0

What runs it

machinebest fitverdictspeed
Apple M1 · 8GB 24B Q3_K_M ❌ Won't fit
Apple M1 Pro · 16GB 24B Q3_K_M 🐢 CPU-only (slow) 1–2 tok/s
Apple M2 · 16GB 24B Q3_K_M 🐢 CPU-only (slow) 0.6–1 tok/s
Apple M3 Pro · 18GB 24B Q4_K_M 🐢 CPU-only (slow) 0.7–1 tok/s
Apple M2 Pro · 32GB 24B Q5_K_M ✅ Runs (tight) 6–10 tok/s
Apple M4 Pro · 48GB 24B Q8_0 ✅ Runs comfortably 6–9 tok/s
Apple M3 Max · 36GB 24B Q5_K_M ✅ Runs (tight) 10–17 tok/s
Apple M2 Max · 64GB 24B Q8_0 ✅ Runs comfortably 7–12 tok/s
Apple M4 Max · 128GB 24B Q8_0 ✅ Runs comfortably 9–15 tok/s
Apple M2 Ultra · 192GB 24B Q8_0 ✅ Runs comfortably 10–16 tok/s
NVIDIA GeForce RTX 3060 · 12GB VRAM 24B Q4_K_M ⚠️ Partial GPU offload 3–5 tok/s
NVIDIA GeForce RTX 4070 · 12GB VRAM 24B Q4_K_M ⚠️ Partial GPU offload 4–7 tok/s
NVIDIA GeForce RTX 4090 · 24GB VRAM 24B Q5_K_M ✅ Runs (tight) 25–40 tok/s
NVIDIA GeForce RTX 5080 · 16GB VRAM 24B Q3_K_M ✅ Runs (tight) 30–50 tok/s
NVIDIA GeForce RTX 5090 · 32GB VRAM 24B Q8_0 ✅ Runs (tight) 30–45 tok/s
AMD Radeon RX 7900 XTX · 24GB VRAM 24B Q5_K_M ✅ Runs (tight) 20–35 tok/s

Shopping for a machine? What hardware do I need to run Mistral Small 3? →

Check your exact rig: open the checker →

$npx runlocal-sh can-i-run mistral-small

Architectural kin

19 models share Mistral Small 3's KV-cache geometry — they size memory the same way, so a rig that fits one tends to fit its same-size kin.

…and 7 more.

Explore the whole catalog by architecture: open the graph →