Chat model by Alibaba · 4.83B parameters
Qwen3 VL 4B Instruct FP8 fits comfortably in your 24 GB Mac using Q8_0 quantization, using 25% of your RAM.
~33 tok/s
Q8_0 · 5.9 GB
| Quantization | Memory | Speed | Fits? |
|---|---|---|---|
| Q8_0 Recommended | 5.9 GB | ~33 tok/s | ✓ |
| Q6_K | 4.7 GB | ~44 tok/s | ✓ |
| Q5_K_M | 4.1 GB | ~51 tok/s | ✓ |
| Q4_K_M | 3.6 GB | ~60 tok/s | ✓ |
| Q3_K_M | 3.1 GB | ~73 tok/s | ✓ |
| Q2_K | 2.6 GB | ~94 tok/s | ✓ |
ToolPiper downloads, manages, and runs models with one click. Apple Silicon optimized.
Get ToolPiper — Free