can it run
No. 32GB of memory can't hold the 307GB working set this model needs. A smaller quant or a bigger machine is required.
| model file | 245GB (GGUF Q4_K_M) |
| minimum memory | 307GB working set |
| this machine | 32GB GDDR7, ~960 GB/s |
| params | Maverick 400B A17B |
| license | llama4 (gated on HF) |
| price | $4500 |
easy run: ollama pull llama4:maverick
runs with: ollama · llama.cpp · koboldcpp · lm studio
→ ASUS ExpertCenter Pro ET900N G3 (GB300) — cheapest machine that runs it ($60000)