6-bit Open on Hugging Face
Parameters
4.0B
Weights (est.)
≈ 3.3 GB
Mac memory (est.)
8 GB+
Size is estimated from the parameter count and quantisation. Memory allows a fifth extra for context plus 2 GB for the system; long contexts need more.
Run it
With mlx-lm on an Apple silicon Mac:
pip install mlx-lmmlx_lm.generate --model mlx-community/Qwen3-4B-6bit --prompt "Hello"
Details
- Base model
- Qwen/Qwen3-4B (quantized) · Qwen3 4B on the radar →
- Vendor
- Qwen
- Task
- Text generation
- Quantisation
- 6-bit
- Licence
- apache-2.0
- Downloads
- 82 in the last 30 days · 6.6K all time · 0 likes
- Published
- 2025-04-28 · updated 2025-04-28 (1y ago)
- Tags
- qwen3
Other MLX builds of Qwen/Qwen3-4B
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| Qwen3-4B-3bit | 3-bit | ≈ 1.8 GB | 8 GB+ | 172 |
| Qwen3-4B-4bit | 4-bit | ≈ 2.3 GB | 8 GB+ | 30K |
| Qwen3-4B-4bit-DWQ | 4-bit DWQ | ≈ 2.3 GB | 8 GB+ | 177 |
| Qwen3-4B-4bit-DWQ-053125 | 4-bit DWQ | ≈ 2.3 GB | 8 GB+ | 151 |
| Qwen3-4B-4bit-DWQ-calibration-1332 | 4-bit DWQ | ≈ 2.3 GB | 8 GB+ | 33 |
| Qwen3-4B-6bit (this one) | 6-bit | ≈ 3.3 GB | 8 GB+ | 82 |
| Qwen3-4B-8bit | 8-bit | ≈ 4.3 GB | 8 GB+ | 1.7K |
| Qwen3-4B-bf16 | bf16 | ≈ 8.0 GB | 16 GB+ | 246 |