MLX Community · First MLX build 17 Mar 2026
Mistral-Small-4-119B-2603
MLX
apache-2.0
A 119B parameter model, with 5 MLX builds, from 4-bit to bf16.
- Parameters
- 119B From Hugging Face
- Quantisation
- 4-bit Mistral-Small-4-119B-2603-4bit
- Memory needed
- 78 GB Estimate, unified memory
- Download
- 67.7 GB Weights, estimated
- Downloads
- 2.3K This build, last 30 days
- Task
- —
- MLX builds
- 5 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 78 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/Mistral-Small-4-119B-2603-4bitMLX builds
5 on mlx-community| Mistral-Small-4-119B-2603-4bit | 4-bit | 67.7 GB | 78 GB | 2.3K |
|---|---|---|---|---|
| Mistral-Small-4-119B-2603-5bit | 5-bit | 82.6 GB | 95 GB | 177 |
| Mistral-Small-4-119B-2603-6bit | 6-bit | 97.5 GB | 111 GB | 174 |
| Mistral-Small-4-119B-2603-8bit | 8-bit | 127 GB | 145 GB | 202 |
| Mistral-Small-4-119B-2603-bf16 | bf16 | 239 GB | 269 GB | 86 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.