Question 1

Can the Jetson Orin NX (16GB) run qwen2.5vl 7B?

Accepted Answer

qwen2.5vl 7B needs about 5.82GB working set (weights 5.6GB + KV cache 0.22GB at 4k context). The Jetson Orin NX (16GB) offers ~14GB usable — verdict: full. 5.8 GB needed of 14 GB usable — headroom for context

Question 2

How fast is qwen2.5vl 7B on the Jetson Orin NX (16GB)?

Accepted Answer

~4.6–10 tok/s on the Jetson Orin NX (16GB). computed roofline: 102.4 GB/s × 0.25–0.55 efficiency window / 5.6GB (Q4_K_M) — bands, never points.

Question 3

What should I run qwen2.5vl 7B on instead?

Accepted Answer

Tesla M60 (16GB, used) runs it fully ($35).

Can the Jetson Orin NX (16GB) run qwen2.5vl 7B?

Run qwen2.5vl 7B on instead