4-bit Open on Hugging Face
Parameters
671B
Weights (est.)
≈ 378 GB
Mac memory (est.)
512 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/DeepSeek-V3.1-4bit --prompt "Hello"
Details
- Base model
- deepseek-ai/DeepSeek-V3.1 (quantized)
- Vendor
- DeepSeek
- Task
- Text generation
- Quantisation
- 4-bit
- Licence
- mit
- Downloads
- 830 in the last 30 days · 106K all time · 6 likes
- Published
- 2025-08-21 · updated 2025-08-21 (1y ago)
- Tags
- deepseek_v3, custom_code
Other MLX builds of deepseek-ai/DeepSeek-V3.1
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| DeepSeek-V3.1-4bit (this one) | 4-bit | ≈ 378 GB | 512 GB+ | 830 |
| DeepSeek-V3.1-mlx-DQ5_K_M | 4-bit | ≈ 378 GB | 512 GB+ | 61 |
| DeepSeek-V3.1-8bit | 8-bit | ≈ 713 GB | > 512 GB | 647 |