MLX Community · First MLX build 3 Dec 2025
Ministral-3-3B-Instruct-2512
MLX
apache-2.0
A 3.8B parameter model, with 5 MLX builds, from 4-bit to bf16.
- Parameters
- 3.8B From Hugging Face
- Quantisation
- 4-bit Ministral-3-3B-Instruct-2512-4bit
- Memory needed
- 6 GB Estimate, unified memory
- Download
- 2.7 GB Weights, estimated
- Downloads
- 20K This build, last 30 days
- Task
- —
- MLX builds
- 5 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 6 GB at 4-bit
8 GB
16 GB
24 GB
32 GB
64 GB
128 GB
Fits with headroomTight, close other appsWon’t fit
Estimated from the size of the weights, plus a fifth for context and working memory and 2 GB for macOS and your apps. Long contexts need more.
Run it
Terminal · Hugging Face CLI$ pip install -U huggingface_hub
$ hf download mlx-community/Ministral-3-3B-Instruct-2512-4bitMLX builds
5 on mlx-community| Ministral-3-3B-Instruct-2512-4bit | 4-bit | 2.7 GB | 6 GB | 20K |
|---|---|---|---|---|
| Ministral-3-3B-Instruct-2512-5bit | 5-bit | 3.2 GB | 6 GB | 64 |
| Ministral-3-3B-Instruct-2512-6bit | 6-bit | 3.6 GB | 7 GB | 71 |
| Ministral-3-3B-Instruct-2512-8bit | 8-bit | 4.5 GB | 7 GB | 210 |
| Ministral-3-3B-Instruct-2512-bf16 | bf16 | 7.7 GB | 11 GB | 115 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.