local intelligence

can it run

Can the NVIDIA B300 (Ultra) 288GB run qwen2.5vl 7B?

FULL

5.8 GB needed of 288 GB usable — headroom for context

288GB · Blackwell Ultra · model file 5.6GB · 7B

~660–1100 tok/s

computed roofline: 8000 GB/s × 0.46–0.76 efficiency window / 5.6GB (Q4_K_M) — bands, never points.

Run qwen2.5vl 7B on instead

Tesla M60 (16GB, used) — cheapest catalog machine that runs qwen2.5vl 7B fully ($35)

NVIDIA DGX Station (GB300) — upgrade path machine

every model the NVIDIA B300 (Ultra) 288GB can run — full list

explore the full catalog90 machines indexed · live prices · what each one can run
2026-09-24 · ← all models on the NVIDIA B300 (Ultra) 288GB · fit = weights + 4k KV vs usable memory