How we got this number
Each generated token reads the active part of the model from memory once: 2 GB. The Minisforum UM790 Pro (Ryzen 9 7940HS, 32GB) moves 89.6 GB/s, so the ceiling is 45 tokens/sec. Real machines reach a fraction of that ceiling; we use 60%, 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 Qwen 3-VL 30B (A3B)
Other models on the Minisforum UM790 Pro (Ryzen 9 7940HS, 32GB)
| lfm 2.5 8B (A1B) | Fast | 44 tok/s |
| Qwen 3 30B (A3B) | Comfortable | 16 tok/s |
| Qwen3 Coder 30B (A3B) | Comfortable | 16 tok/s |
| Qwen3.5 35B (A3B) | Comfortable | 14 tok/s |
| Qwen3.6 35B (A3B) | Comfortable | 14 tok/s |
| orca mini 7B | Comfortable | 13 tok/s |