VLM Inference Benchmark Explorer
Cosmos-Reason1-7B on Jetson AGX Thor — BF16, vLLM vision-language inference benchmark
Open this configuration in the VLM Inference Benchmark Explorer
| Model | Cosmos-Reason1-7B |
|---|---|
| Parameters | 8.3B |
| Device | Jetson AGX Thor |
| 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 | 582 | 16.4 | ≤3000 ms |
| 2 | 764 | 15.9 | ≤3000 ms |
| 4 | 1315 | 15.3 | ≤3000 ms |
| 8 | 1896 | 13.6 | ≤3000 ms |
| 16 | 3332 | 11.9 | >3000 ms |
| Images per request | TTFT (ms) | TPS (tok/s) | Target |
|---|---|---|---|
| 1 | 582 | 16.4 | ≤3000 ms |
| 2 | 1012 | 16.2 | ≤3000 ms |
| 4 | 1975 | 16.1 | ≤3000 ms |
| 8 | 4095 | 16.0 | >3000 ms |
| 16 | 8918 | 15.2 | >3000 ms |
| 32 | 21320 | 14.5 | >3000 ms |
| 64 | 56100 | 12.8 | >3000 ms |
| 128 | — | — | Failed |
Cosmos-Reason1-7B (8.3B parameters), served in BF16 format with vLLM on Jetson AGX Thor, starts answering a single camera that sends one 720p image after 582 ms (time to first token, TTFT), then writes 16.4 tok/s.
With 64 images per camera at 720p, the answer starts after 56,100 ms. With one 2K image, after 2,297 ms.
At the default target of a mean TTFT within 3,000 ms (720p, one image per camera), it keeps up with 8 cameras at once.
Cosmos-Reason1-7B (8.3B parameters), served in BF16 format with vLLM on Jetson AGX Thor, starts answering a single camera that sends one 720p image after 582 ms (time to first token, TTFT), then writes 16.4 tok/s.
With 64 images per camera at 720p, the answer starts after 56,100 ms. With one 2K image, after 2,297 ms.
At the default target of a mean TTFT within 3,000 ms (720p, one image per camera), it keeps up with 8 cameras at once.