General model by chwan · 26.62B parameters
Everything this chip runs, machine by machine: MacBook Pro 14" M5 Pro, MacBook Pro 16" M5 Pro
DeepSeek V3 5layer can run but will be tight at 85% of RAM. Consider a lower quantization or closing other apps.
— tok/s
Q5_K_M · 20.3 GB · Benchmark needed
Alternative: Q4_K_M (17.7 GB, ~137 tok/s)
| Quantization | Memory | Speed | Fits? |
|---|---|---|---|
| Q8_0 | 30.2 GB | ~76 tok/s | ✗ |
| Q6_K | 23.5 GB | ~99 tok/s | ✗ |
| Q5_K_M Recommended | 20.3 GB | ~117 tok/s | ✓ |
| Q4_K_M | 17.7 GB | ~137 tok/s | ✓ |
| Q3_K_M | 15.0 GB | ~166 tok/s | ✓ |
| Q2_K | 12.1 GB | ~215 tok/s | ✓ |
Closest matches by use case and parameter count.
ToolPiper downloads, manages, and runs models with one click. Apple Silicon optimized.
Get ToolPiper — Free