MiniCPM-V-4.6 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	MiniCPM-V-4.6
Parameters	1.3B
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	1.12	73.2	≤3 s
2	1.89	29.8	≤3 s
4	3.66	47.6	>3 s
8	8.80	49.1	>3 s

Images per request	Response (s)	TPS (tok/s)	Target
1	1.12	73.2	≤3 s
2	2.18	73.3	≤3 s
4	4.18	72.4	>3 s
8	8.25	72.4	>3 s
16	16.4	69.1	>3 s
32	33.2	63.2	>3 s

MiniCPM-V-4.6 (1.3B 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 1.12 s, then writes 73.2 tok/s.

With 32 images per camera at 720p, the answer starts after 33.2 s. With one 2K image, after 1.54 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.