VPS Arena

Run Mistral Small 3.2 24B on a VPS.

September 2026 · 68 of 280 plans fit · re-ranked daily

Dense 24B with vision and function calling; the practical ceiling for a 24 GB box. On a CPU-only VPS it needs about 17.1 GB of RAM: 14.3 GB of Q4_K_M weights, 1.3 GB of KV cache for an 8,192-token context and 1.5 GB for the OS and runtime. 68 of the 280 plans in our index fit; the cheapest comfortable pick is Contabo's Cloud VPS 8 · 8 vCPU · 24 GB at $16.27/mo, streaming an estimated 1.3–2.6 tok/s.

RAM needed · CPU inference

17.1 GB

weights · Q4_K_M
14.3 GB
KV cache · 8K context
1.3 GB
OS + runtime headroom
1.5 GB

Other quants: Q8_0 weights 25.1 GB (near-lossless, about half of F16); F16 47.2 GB. Comfortable = 20% headroom over the total.

Model card

Size
24B parameters
Context
128K tokens
Kind
chat · vision
Released
2025-06
License
Apache-2.0

Facts fetched from Hugging Face on 3 Sept 2026: exact GGUF file sizes, KV geometry from the GGUF header · 156,348 downloads.

ollama run mistral-small3.2:24bmodel card ↗Mistral AI

Best VPS plans for Mistral Small 3.2 24B

ranked by estimated tokens/s per dollar, comfortable fits first

  1. 1
    ContaboCloud VPS 8 · 8 vCPU · 24 GB

    8 vCPU shared · 24 GB RAM · 300 GB SSD

    runs comfortably~1.3–2.6 tok/s$16.27/mo
  2. 2
    netcupRS 4000 G12 · 12 dedicated cores · 32 GB

    12 vCPU dedicated · 32 GB RAM · 1024 GB NVME

    runs comfortably~2.5–4.9 tok/s$38.99/mo
  3. 3
    ContaboCloud VPS 12 · 12 vCPU · 48 GB

    12 vCPU shared · 48 GB RAM · 400 GB SSD

    runs comfortably~1.6–3.3 tok/s$29.05/mo
  4. 4
    netcupVPS 4000 G12 · 12 vCore · 32 GB

    12 vCPU shared · 32 GB RAM · 1024 GB NVME

    runs comfortably~1.6–3.3 tok/s$31.66/mo
  5. 5
    OVHcloudVPS-4 2027 · 8 vCPU / 24 GB

    8 vCPU shared · 24 GB RAM · 200 GB NVME

    runs comfortably~1.3–2.6 tok/s$27.50/mo
  6. 6
    HetznerCX53 · 16 vCPU · 32 GB

    16 vCPU shared · 32 GB RAM · 320 GB NVME

    runs comfortably~1.6–3.3 tok/s$34.27/mo

Speed = effective memory bandwidth ÷ active weight bytes (4 GB/s per shared vCPU, 6 per dedicated), shown as a band. Real numbers depend on the host CPU generation, AVX-512/AMX support and how noisy the neighbours are — treat these as order-of-magnitude.

Or buy hardware · 9 reference machines fit

all machines →
  • Mac mini (M6, 32 GB)32 GB · 153 GB/s · ~5.5–8.5 tok/s$1,299= 80 mo of VPS
  • Mac mini (M5 Pro, 24 GB)24 GB · 307 GB/s · ~11–17 tok/s$1,699= 104 mo of VPS
  • Framework Desktop (Ryzen AI Max+ 395, 64 GB)64 GB · 256 GB/s · ~9.2–14 tok/s$1,959= 10+ yrs of VPS
  • Mac mini (M5 Pro, 64 GB)64 GB · 307 GB/s · ~11–17 tok/s$2,299= 10+ yrs of VPS
  • Mac Studio (M5 Max, 36 GB)36 GB · 460 GB/s · ~17–25 tok/s$2,499= 10+ yrs of VPS
  • Framework Desktop (Ryzen AI Max+ 395, 128 GB)128 GB · 256 GB/s · ~9.2–14 tok/s$3,449= 10+ yrs of VPS

Buy · per month

$37.45

$36.08 hardware + $1.36 power

Rent · per month

$16.27

Contabo Cloud VPS 8 · 8 vCPU · 24 GB

Break-even

The Mac mini (M6, 32 GB) pays for itself after 87 months of replacing the VPS — and it streams an estimated 5.5–8.5 tok/s against the VPS's CPU-only pace.

* approximate: August 2026 US retail median rather than list price. Local speed = peak bandwidth × 0.7 efficiency (0.35 on CPU-only boards) ÷ active weight bytes; unified-memory machines are assumed to give models 75% of their RAM. GPU cards need a host PC that is not included in the price.

Frequently asked

How much RAM does Mistral Small 3.2 24B need?
About 17.1 GB for CPU inference at Q4_K_M: 14.3 GB of weights, 1.3 GB of KV cache at 8,192 tokens of context, and 1.5 GB of headroom. Longer contexts need more KV cache (160 KB per token for this model).
What is the cheapest VPS that can run Mistral Small 3.2 24B?
Contabo Cloud VPS 8 · 8 vCPU · 24 GB (24 GB RAM, 8 vCPU) at $16.27/mo excl. VAT runs it comfortably as of 5 Sept 2026.
How fast will Mistral Small 3.2 24B run on a VPS without a GPU?
Roughly 1.3–2.6 tok/s on the top pick. Token generation is bound by memory bandwidth — each token streams the full weights once — so more and dedicated vCPUs help; a GPU or Apple-silicon machine is 10–50× faster.
Same lineupMinistral 8BMistral Nemo 12BDevstral Small 24BSame size classQwen3 32BQwen2.5-Coder 32BGemma 3 27BDeepSeek-R1 Distill Qwen 32BVPS for AI agents (API-hosted models) →