can it run
No. 11GB of memory can't hold the 28GB working set this model needs. A smaller quant or a bigger machine is required.
weights 22GB + KV cache 1GB at 4k context vs ~11GB usable → no. 23.0 GB vs 11 GB usable — does not fit
computed roofline: 484 GB/s × 0.28–0.50 efficiency window / 22.0GB (Q4_K_M) — bands, never points.
| model file | 22GB (GGUF Q4_K_M) |
| minimum memory | 28GB working set |
| this machine | 11GB GDDR5X, 484 GB/s |
| params | 32B |
| license | apache-2.0 |
| price | $150 |
easy run: ollama pull qwen2.5vl:32b
runs with: ollama · llama.cpp · koboldcpp · lm studio
→ Mac mini M5 Pro (64GB) — cheapest machine that runs it ($1669)
→ every model the GTX 1080 Ti (11GB, used) can run — full list
→ RTX 3090 (24GB, used) — upgrade path machine