Compatibility check
Can I run Qwen2.5 VL 72B Instruct on RTX PRO 6000 Blackwell?
Yes — it runs.best quant: Q4_K_M
Decode speed
32 tok/sest
Usable context
66K
Memory
96 GB
Q4_K_MFits in memory32 tok/sest · 66K ctx
Q5_K_MFits in memory28 tok/sest · 66K ctx
Q8_0Fits in memory18 tok/sest · 16K ctx