VLM Inference Benchmark Explorer
Qwen3-VL-30B-A3B-Instruct on RTX PRO 6000 — FP8, vLLM vision-language inference benchmark
Open this configuration in the VLM Inference Benchmark Explorer
| Model | Qwen3-VL-30B-A3B-Instruct |
|---|---|
| Parameters | 30B |
| Device | RTX PRO 6000 |
| Quantization | FP8 |
| Image size | 720p |
| Images per camera | 1 |
| Max Cameras (TTFT ≤ 3000 ms) | 16+ |
| Engine | vLLM |
| Cameras | TTFT (ms) | TPS (tok/s) | Target |
|---|---|---|---|
| 1 | 161 | 194.4 | ≤3000 ms |
| 2 | 257 | 153.3 | ≤3000 ms |
| 4 | 324 | 127.4 | ≤3000 ms |
| 8 | 425 | 100.6 | ≤3000 ms |
| 16 | 550 | 82.3 | ≤3000 ms |
| Images per request | TTFT (ms) | TPS (tok/s) | Target |
|---|---|---|---|
| 1 | 161 | 194.4 | ≤3000 ms |
| 3 | 373 | 186.1 | ≤3000 ms |
| 5 | 1041 | 182.7 | ≤3000 ms |
Qwen3-VL-30B-A3B-Instruct (30B parameters), served in FP8 format with vLLM on RTX PRO 6000, starts answering a single camera that sends one 720p image after 161 ms (time to first token, TTFT), then writes 194.4 tok/s.
With 5 images per camera at 720p, the answer starts after 1,041 ms. With one 2K image, after 581 ms.
At the default target of a mean TTFT within 3,000 ms (720p, one image per camera), it keeps up with at least 16 cameras at once: it still met the target at the highest number measured, so the real limit is higher.
Qwen3-VL-30B-A3B-Instruct (30B parameters), served in FP8 format with vLLM on RTX PRO 6000, starts answering a single camera that sends one 720p image after 161 ms (time to first token, TTFT), then writes 194.4 tok/s.
With 5 images per camera at 720p, the answer starts after 1,041 ms. With one 2K image, after 581 ms.
At the default target of a mean TTFT within 3,000 ms (720p, one image per camera), it keeps up with at least 16 cameras at once: it still met the target at the highest number measured, so the real limit is higher.