Chat model by redhatai · 19.6B parameters
Llama 4 Scout 17B 16E Instruct quantized.w4a16 fits comfortably in your 24 GB Mac using Q6_K quantization, using 73% of your RAM.
~98 tok/s
Q6_K · 17.5 GB
| Quantization | Memory | Speed | Fits? |
|---|---|---|---|
| Q8_0 | 22.4 GB | ~75 tok/s | ✗ |
| Q6_K Recommended | 17.5 GB | ~98 tok/s | ✓ |
| Q5_K_M | 15.1 GB | ~116 tok/s | ✓ |
| Q4_K_M | 13.2 GB | ~136 tok/s | ✓ |
| Q3_K_M | 11.2 GB | ~164 tok/s | ✓ |
| Q2_K | 9.0 GB | ~213 tok/s | ✓ |
ToolPiper downloads, manages, and runs models with one click. Apple Silicon optimized.
Get ToolPiper — Free