VLM Inference Benchmark Explorer
Cosmos3-Edge on Jetson AGX Orin 32GB — BF16, vLLM vision-language inference benchmark
Open this configuration in the VLM Inference Benchmark Explorer
| Model | Cosmos3-Edge |
|---|---|
| Parameters | 3.9B |
| Device | Jetson AGX Orin 32GB |
| Quantization | BF16 |
| Image size | 720p |
| Images per camera | 1 |
| Max Cameras (TTFT ≤ 3000 ms) | 8+ |
| Engine | vLLM |
| Cameras | TTFT (ms) | TPS (tok/s) | Target |
|---|---|---|---|
| 1 | 684 | 42.6 | ≤3000 ms |
| 2 | 802 | 35.6 | ≤3000 ms |
| 4 | 1602 | 32.0 | ≤3000 ms |
| 8 | 2887 | 26.2 | ≤3000 ms |
| Images per request | TTFT (ms) | TPS (tok/s) | Target |
|---|---|---|---|
| 1 | 684 | 42.6 | ≤3000 ms |
| 3 | 1755 | 40.5 | ≤3000 ms |
| 5 | 2956 | 38.5 | ≤3000 ms |
Cosmos3-Edge (3.9B parameters), served in BF16 format with vLLM on Jetson AGX Orin 32GB, starts answering a single camera that sends one 720p image after 684 ms (time to first token, TTFT), then writes 42.6 tok/s.
With 5 images per camera at 720p, the answer starts after 2,956 ms. With one 2K image, after 3,951 ms.
At the default target of a mean TTFT within 3,000 ms (720p, one image per camera), it keeps up with at least 8 cameras at once: it still met the target at the highest number measured, so the real limit is higher.
Cosmos3-Edge (3.9B parameters), served in BF16 format with vLLM on Jetson AGX Orin 32GB, starts answering a single camera that sends one 720p image after 684 ms (time to first token, TTFT), then writes 42.6 tok/s.
With 5 images per camera at 720p, the answer starts after 2,956 ms. With one 2K image, after 3,951 ms.
At the default target of a mean TTFT within 3,000 ms (720p, one image per camera), it keeps up with at least 8 cameras at once: it still met the target at the highest number measured, so the real limit is higher.