General model by opengvlab · 4.15B parameters
InternVL2 4B fits comfortably in your 24 GB Mac using Q8_0 quantization, using 21% of your RAM.
~39 tok/s
Q8_0 · 5.1 GB
| Quantization | Memory | Speed | Fits? |
|---|---|---|---|
| Q8_0 Recommended | 5.1 GB | ~39 tok/s | ✓ |
| Q6_K | 4.1 GB | ~51 tok/s | ✓ |
| Q5_K_M | 3.6 GB | ~60 tok/s | ✓ |
| Q4_K_M | 3.2 GB | ~70 tok/s | ✓ |
| Q3_K_M | 2.8 GB | ~85 tok/s | ✓ |
| Q2_K | 2.3 GB | ~110 tok/s | ✓ |
ToolPiper downloads, manages, and runs models with one click. Apple Silicon optimized.
Get ToolPiper — Free