Apple M4 Pro meets the minimum VRAM requirement for Q4 inference of inference-net/Schematron-3B. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
Apple M4 Pro can run inference-net/Schematron-3B with Q4 quantization. At approximately 41 tokens/second, you can expect Moderate speed - useful for batch processing.
You have 62GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 2GB | 64GB | 40.64 tok/s | ✅ Fits comfortably |
| Q8 | 3GB | 64GB | 31.80 tok/s | ✅ Fits comfortably |
| FP16 | 6GB | 64GB | 17.32 tok/s | ✅ Fits comfortably |