4-bit Open on Hugging Face
Parameters
3.1B
Weights (est.)
≈ 1.7 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/SmolLM3-3B-4bit --prompt "Hello"
Details
- Base model
- HuggingFaceTB/SmolLM3-3B (quantized) · SmolLM3 3B on the radar →
- Vendor
- Hugging Face
- Task
- Text generation
- Quantisation
- 4-bit
- Licence
- apache-2.0
- Downloads
- 3.5K in the last 30 days · 48K all time · 7 likes
- Published
- 2025-07-08 · updated 2025-07-08 (1y ago)
- Languages
- en, fr, es, it, pt, zh, ar, ru
- Tags
- smollm3
Other MLX builds of HuggingFaceTB/SmolLM3-3B
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| SmolLM3-3B-3bit | 3-bit | ≈ 1.3 GB | 8 GB+ | 59 |
| SmolLM3-3B-4bit (this one) | 4-bit | ≈ 1.7 GB | 8 GB+ | 3.5K |
| SmolLM3-3B-4bit-DWQ | 4-bit DWQ | ≈ 1.7 GB | 8 GB+ | 74 |
| SmolLM3-3B-5bit | 5-bit | ≈ 2.1 GB | 8 GB+ | 48 |
| SmolLM3-3B-6bit | 6-bit | ≈ 2.5 GB | 8 GB+ | 35 |
| SmolLM3-3B-8bit | 8-bit | ≈ 3.3 GB | 8 GB+ | 131 |
| SmolLM3-3B-bf16 | bf16 | ≈ 6.2 GB | 16 GB+ | 62 |
| SmolLM3-3B-abliterated-bf16 | bf16 | ≈ 6.2 GB | 16 GB+ | 54 |
| SmolLM3-3B-gabliterated-bf16 | bf16 | ≈ 6.2 GB | 16 GB+ | 30 |