VLM Inference Benchmark Explorer
Qwen3-VL-4B-Instruct on Jetson AGX Orin 32GB — Q8_0, llama.cpp vision-language inference benchmark
Open this configuration in the VLM Inference Benchmark Explorer
| Model | Qwen3-VL-4B-Instruct |
|---|---|
| Parameters | 4.4B |
| Device | Jetson AGX Orin 32GB |
| 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) | Status |
|---|---|---|---|
| 1 | 1.31 | 28.4 | PASS |
| 2 | 1.96 | 19.7 | PASS |
| 4 | 3.98 | 10.2 | FAIL |
| 8 | 6.61 | 14.3 | FAIL |
Qwen3-VL-4B-Instruct (4.4B parameters), served in Q8_0 format with llama.cpp on Jetson AGX Orin 32GB, starts answering a single camera that sends one 720p image after 1.31 s, then writes 28.4 tok/s.
With 5 images per camera at 720p, the answer starts after 6.4 s. With one 2K image, after 8.32 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.
Qwen3-VL-4B-Instruct (4.4B parameters), served in Q8_0 format with llama.cpp on Jetson AGX Orin 32GB, starts answering a single camera that sends one 720p image after 1.31 s, then writes 28.4 tok/s.
With 5 images per camera at 720p, the answer starts after 6.4 s. With one 2K image, after 8.32 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.