General model by applied-innovation-center · 40.67B parameters
Everything this chip runs, machine by machine: MacBook Pro 14" M5 Pro, MacBook Pro 16" M5 Pro
Karnak 40B v1.0 can run but will be tight at 76% of RAM. Consider a lower quantization or closing other apps.
— tok/s
Q2_K · 18.2 GB · Benchmark needed
| Quantization | Memory | Speed | Fits? |
|---|---|---|---|
| Q8_0 | 45.9 GB | ~36 tok/s | ✗ |
| Q6_K | 35.7 GB | ~47 tok/s | ✗ |
| Q5_K_M | 30.8 GB | ~56 tok/s | ✗ |
| Q4_K_M | 26.8 GB | ~65 tok/s | ✗ |
| Q3_K_M | 22.7 GB | ~79 tok/s | ✗ |
| Q2_K Recommended | 18.2 GB | ~103 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