General model by opengvlab · 30.85B parameters
InternVL3_5 30B A3B can run but will be tight at 85% of RAM. Consider a lower quantization or closing other apps.
~9 tok/s
Q4_K_M · 20.4 GB
Alternative: Q3_K_M (17.3 GB, ~11 tok/s)
| Quantization | Memory | Speed | Fits? |
|---|---|---|---|
| Q8_0 | 34.9 GB | ~5 tok/s | ✗ |
| Q6_K | 27.2 GB | ~7 tok/s | ✗ |
| Q5_K_M | 23.5 GB | ~8 tok/s | ✗ |
| Q4_K_M Recommended | 20.4 GB | ~9 tok/s | ✓ |
| Q3_K_M | 17.3 GB | ~11 tok/s | ✓ |
| Q2_K | 13.9 GB | ~15 tok/s | ✓ |
GGUF Sources
ToolPiper downloads, manages, and runs models with one click. Apple Silicon optimized.
Get ToolPiper — Free