How we got this number
Each generated token reads the whole model from memory once: 8 GB. The NVIDIA DGX Spark (128GB) moves 273 GB/s, so the ceiling is 34 tokens/sec. Real machines reach a fraction of that ceiling; we use 70%, an assumption until we get measured runs for this kind of machine. Send us your real numbers and we'll replace the estimate.
Faster machines for LLaVA 13B
Other models on the NVIDIA DGX Spark (128GB)
| lfm 2.5 8B (A1B) | Fast | 120+ tok/s |
| Gemma 2B | Fast | 112 tok/s |
| smollm 2 1.7B | Fast | 106 tok/s |
| starCoder 2 3B | Fast | 106 tok/s |
| falcon 3 1B | Fast | 106 tok/s |
| Qwen2.5 3B | Fast | 101 tok/s |