VLM Inference Benchmark Explorer
Cosmos-Reason2-8B on Jetson AGX Orin 32GB Developer Kit — Q4_K_M, llama.cpp vision-language inference benchmark
Open this configuration in the VLM Inference Benchmark Explorer
| Model | Cosmos-Reason2-8B |
|---|---|
| Parameters | 8.8B |
| Device | Jetson AGX Orin 32GB Developer Kit |
| Quantization | Q4_K_M |
| Image size | 720p |
| Images per camera | 1 |
| Max Cameras (response ≤ 3 s) | 1 |
| Engine | llama.cpp |
| Cameras | Response (s) | TPS (tok/s) | Target |
|---|---|---|---|
| 1 | 2.71 | 25.4 | ≤3 s |
| 2 | 3.17 | 10.6 | >3 s |
| 4 | 5.13 | 10.0 | >3 s |
| 8 | 15.5 | 12.0 | >3 s |
| Images per request | Response (s) | TPS (tok/s) | Target |
|---|---|---|---|
| 1 | 2.71 | 25.4 | ≤3 s |
| 3 | 7.93 | 24.3 | >3 s |
| 5 | 13.2 | 23.5 | >3 s |
Cosmos-Reason2-8B (8.8B parameters), served in Q4_K_M format with llama.cpp on Jetson AGX Orin 32GB Developer Kit, starts answering a single camera that sends one 720p image after 2.71 s, then writes 25.4 tok/s.
With 5 images per camera at 720p, the answer starts after 13.21 s. With one 2K image, after 20.51 s.
At the default target of an answer starting within 3 s (720p, one image per camera), it keeps up with one camera at a time.
Cosmos-Reason2-8B (8.8B parameters), served in Q4_K_M format with llama.cpp on Jetson AGX Orin 32GB Developer Kit, starts answering a single camera that sends one 720p image after 2.71 s, then writes 25.4 tok/s.
With 5 images per camera at 720p, the answer starts after 13.21 s. With one 2K image, after 20.51 s.
At the default target of an answer starting within 3 s (720p, one image per camera), it keeps up with one camera at a time.