Can I Run Mistral Small 4 on a NVIDIA DGX Station (Blackwell Ultra)?
Runs at full precision (f32). Zero quality loss.
3 quantizations fit your 784GB
| Quant | Min VRAM | Recommended | File size | Headroom |
|---|---|---|---|---|
| f32BEST | 478.6 GB | 480.1 GB | 1.7 GB | +305.4 GB |
| fp16 | 239.8 GB | 241.3 GB | 0.9 GB | +544.2 GB |
| Q3_K_S | 47.0 GB | 48.5 GB | 49.6 GB | +737.0 GB |
Try it in the cloud first
Don't want to download Mistral Small 4 just to try it? Use a hosted API or rent a GPU by the second.
Affiliate links — we earn a commission at no cost to you.
All quant variants, benchmark scores, and use-case tags.
Top-ranked open-source models that fit in 784GB.
FAQ
Can the NVIDIA DGX Station (Blackwell Ultra) run Mistral Small 4?
Yes. The NVIDIA DGX Station (Blackwell Ultra)'s 784GB of unified memory is enough to run Mistral Small 4 at f32 quantization (478.6GB required).
What's the best quantization to use?
f32 is the highest-precision quantization that fits in your 784GB. It uses about 478.6GB of memory and 480.1GB recommended for comfortable inference.
What if I need more headroom for context length?
KV cache memory grows with context length. The numbers above assume a baseline 2K-4K context. For long-context use (32K+), add another 2-6GB depending on the model architecture.