Qwen3.5-0.8B on Jetson AGX Orin 32GB Developer Kit — Q8_0, llama.cpp vision-language inference benchmark

Open this configuration in the VLM Inference Benchmark Explorer

Model	Qwen3.5-0.8B
Parameters	0.9B
Device	Jetson AGX Orin 32GB Developer Kit
Quantization	Q8_0
Image size	720p
Images per camera	1
Max Cameras (response ≤ 3 s)	4
Engine	llama.cpp

Cameras	Response (s)	TPS (tok/s)	Target
1	0.53	76.9	≤3 s
2	0.78	49.8	≤3 s
4	1.51	49.6	≤3 s
8	4.86	44.9	>3 s

Images per request	Response (s)	TPS (tok/s)	Target
1	0.53	76.9	≤3 s
2	0.98	77.5	≤3 s
4	1.84	75.5	≤3 s
8	3.58	71.2	>3 s
16	7.18	66.2	>3 s
32	14.9	60.0	>3 s

Qwen3.5-0.8B (0.9B parameters), served in Q8_0 format with llama.cpp on Jetson AGX Orin 32GB Developer Kit, starts answering a single camera that sends one 720p image after 0.53 s, then writes 76.9 tok/s.

With 32 images per camera at 720p, the answer starts after 14.85 s. With one 2K image, after 3.06 s.

At the default target of an answer starting within 3 s (720p, one image per camera), it keeps up with 4 cameras at once.