Qwen-Drive-1.0-4B 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	Qwen-Drive-1.0-4B
Parameters	4.5B
Device	Jetson AGX Orin 32GB Developer Kit
Quantization	Q8_0
Image size	720p
Images per camera	1
Max Cameras (response ≤ 3 s)	2
Engine	llama.cpp

Cameras	Response (s)	TPS (tok/s)	Target
1	1.49	24.0	≤3 s
2	2.83	26.2	≤3 s
4	5.47	19.9	>3 s
8	15.3	20.0	>3 s

Images per request	Response (s)	TPS (tok/s)	Target
1	1.49	24.0	≤3 s
2	2.73	23.9	≤3 s
4	5.26	23.7	>3 s
8	10.3	22.8	>3 s
16	20.8	21.7	>3 s
32	42.9	19.7	>3 s

Qwen-Drive-1.0-4B (4.5B 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 1.49 s, then writes 24 tok/s.

With 32 images per camera at 720p, the answer starts after 42.94 s. With one 2K image, after 8.68 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.