General model by llm-jp · 32.14B parameters
Everything this chip runs, machine by machine: MacBook Pro 14" M5 Pro, MacBook Pro 16" M5 Pro
llm jp 4 32b a3b thinking can run but will be tight at 75% of RAM. Consider a lower quantization or closing other apps.
— tok/s
Q3_K_M · 18.0 GB · Benchmark needed
Alternative: Q2_K (14.5 GB, ~130 tok/s)
| Quantization | Memory | Speed | Fits? |
|---|---|---|---|
| Q8_0 | 36.4 GB | ~46 tok/s | ✗ |
| Q6_K | 28.3 GB | ~60 tok/s | ✗ |
| Q5_K_M | 24.5 GB | ~71 tok/s | ✗ |
| Q4_K_M | 21.2 GB | ~83 tok/s | ✗ |
| Q3_K_M Recommended | 18.0 GB | ~100 tok/s | ✓ |
| Q2_K | 14.5 GB | ~130 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