BAAI · First MLX build 16 Mar 2026
bge-m3
MLX
Text mit
A 568M parameter embeddings model from BAAI, with 4 MLX builds, from 4-bit to fp16.
- Parameters
- 568M From Hugging Face
- Quantisation
- fp16 bge-m3-mlx-fp16
- Memory needed
- 4 GB Estimate, unified memory
- Download
- 1.1 GB Weights, estimated
- Downloads
- 3.2K This build, last 30 days
- Task
- Embeddings Text
- MLX builds
- 4 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 4 GB at fp16
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/bge-m3-mlx-fp16MLX builds
4 on mlx-community| bge-m3-mlx-4bit | 4-bit | 320 MB | 3 GB | 575 |
|---|---|---|---|---|
| bge-m3-mlx-6bit | 6-bit | 462 MB | 3 GB | 155 |
| bge-m3-mlx-8bit | 8-bit | 604 MB | 3 GB | 1.8K |
| bge-m3-mlx-fp16 | fp16 | 1.1 GB | 4 GB | 3.2K |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.