How we got this number
Each generated token reads the whole model from memory once: 15 GB. The Mac Studio M5 Max (40-core GPU, 64GB) moves 614 GB/s, so the ceiling is 41 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 Mistral small3.2 24B
| Mac Studio M5 Ultra (96GB) | 66 tok/s | Check price → |
| Mac Studio M5 Ultra (256GB) | 66 tok/s | Check price → |
| Mac mini M5 Pro (64GB) | 16 tok/s | Check price → |
| NVIDIA DGX Spark (128GB) | 13 tok/s | Check price → |
Other models on the Mac Studio M5 Max (40-core GPU, 64GB)
| lfm 2.5 8B (A1B) | Fast | 120+ tok/s |
| Gemma 3 4B | Fast | 120+ tok/s |
| Qwen 3 30B (A3B) | Fast | 120+ tok/s |
| Qwen3.5 4B | Fast | 120+ tok/s |
| Qwen3.5 35B (A3B) | Fast | 120+ tok/s |
| Qwen3 Coder 30B (A3B) | Fast | 120+ tok/s |