Can I Run Mistral Medium 3.5 128B on a NVIDIA DGX Station (Blackwell Ultra)?
Runs at full precision (f32). Zero quality loss.
6 quantizations fit your 784GB
| Quant | Min VRAM | Recommended | File size | Headroom |
|---|---|---|---|---|
| f32BEST | 511.8 GB | 513.3 GB | 10.7 GB | +272.2 GB |
| fp16 | 256.4 GB | 257.9 GB | 5.4 GB | +527.6 GB |
| Q6_K | 106.2 GB | 107.7 GB | 102.6 GB | +677.8 GB |
| Q5_K_M | 91.7 GB | 93.2 GB | 88.3 GB | +692.3 GB |
| Q4_K_M | 78.4 GB | 79.9 GB | 74.9 GB | +705.6 GB |
| Q2_K | 43.0 GB | 44.5 GB | 46.6 GB | +741.0 GB |
Try it in the cloud first
Don't want to download Mistral Medium 3.5 128B 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 Medium 3.5 128B?
Yes. The NVIDIA DGX Station (Blackwell Ultra)'s 784GB of unified memory is enough to run Mistral Medium 3.5 128B at f32 quantization (511.8GB required).
What's the best quantization to use?
f32 is the highest-precision quantization that fits in your 784GB. It uses about 511.8GB of memory and 513.3GB 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.