Qwen3.5-4B on Jetson AGX Orin 32GB Developer Kit — Q4_K_M, llama.cpp vision-language inference benchmark

Open this configuration in the VLM Inference Benchmark Explorer

Model	Qwen3.5-4B
Parameters	4.7B
Device	Jetson AGX Orin 32GB Developer Kit
Quantization	Q4_K_M
Image size	720p
Images per camera	1
Max Cameras (response ≤ 3 s)	2
Engine	llama.cpp

Cameras	Response (s)	TPS (tok/s)	Target
1	1.57	33.6	≤3 s
2	2.50	18.2	≤3 s
4	5.16	11.0	>3 s
8	10.1	16.0	>3 s

Images per request	Response (s)	TPS (tok/s)	Target
1	1.57	33.6	≤3 s
3	4.34	33.2	>3 s
5	7.14	33.0	>3 s

Qwen3.5-4B (4.7B parameters), served in Q4_K_M format with llama.cpp on Jetson AGX Orin 32GB Developer Kit, starts answering a single camera that sends one 720p image after 1.57 s, then writes 33.6 tok/s.

With 5 images per camera at 720p, the answer starts after 7.14 s. With one 2K image, after 9.04 s.

At the default target of an answer starting within 3 s (720p, one image per camera), it keeps up with 2 cameras at once.