General model by baidu · 21.95B parameters
Everything this chip runs, machine by machine: MacBook Pro 14" M5 Pro, MacBook Pro 16" M5 Pro
ERNIE 4.5 21B A3B PT fits comfortably in your 24 GB Mac using Q6_K quantization, using 81% of your RAM.
~10 tok/s
Q6_K · 19.5 GB · Estimated
| Quantization | Memory | Speed | Fits? |
|---|---|---|---|
| Q8_0 | 25.0 GB | ~7 tok/s | ✗ |
| Q6_K Recommended | 19.5 GB | ~10 tok/s | ✓ |
| Q5_K_M | 16.9 GB | ~11 tok/s | ✓ |
| Q4_K_M | 14.7 GB | ~13 tok/s | ✓ |
| Q3_K_M | 12.5 GB | ~16 tok/s | ✓ |
| Q2_K | 10.1 GB | ~21 tok/s | ✓ |
GGUF Sources
Closest matches by use case and parameter count.
ToolPiper downloads, manages, and runs models with one click. Apple Silicon optimized.
Get ToolPiper — Free