How we got this number
Each generated token reads the whole model from memory once: 2.3 GB. The Mac mini M6 (16GB) moves 170 GB/s, so the ceiling is 74 tokens/sec. Real machines reach a fraction of that ceiling; we use 80%, calibrated against a measured Mac run. Send us your real numbers and we'll replace the estimate.
Faster machines for Qwen 4B
Other models on the Mac mini M6 (16GB)
| DeepSeek-R1 1.5B | Fast | 120+ tok/s |
| Qwen3.5 0.8B | Fast | 120+ tok/s |
| Qwen 1.8B | Fast | 120+ tok/s |
| Llama 3.2 1B | Fast | 105 tok/s |
| Qwen 3 1.7B | Fast | 97 tok/s |
| Granite 3.1 moe 1B | Fast | 97 tok/s |