MLX Community · First MLX build 30 Jul 2025
Qwen3-30B-A3B-Instruct-2507
MLX
A 31B parameter model, with 3 MLX builds, from 4-bit DWQ to 8-bit DWQ.
- Parameters
- 31B From Hugging Face
- Quantisation
- 4-bit DWQ Qwen3-30B-A3B-Instruct-2507-4bit-DWQ
- Memory needed
- 9 GB Estimate, unified memory
- Download
- 6.0 GB Weights, estimated
- Downloads
- 52 This build, last 30 days
- Task
- —
- MLX builds
- 3 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 9 GB at 4-bit DWQ
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/Qwen3-30B-A3B-Instruct-2507-4bit-DWQMLX builds
3 on mlx-community| Qwen3-30B-A3B-Instruct-2507-4bit-DWQ | 4-bit DWQ | 6.0 GB | 9 GB | 52 |
|---|---|---|---|---|
| Qwen3-30B-A3B-Instruct-2507-6bit-DWQ | 6-bit DWQ | 8.5 GB | 12 GB | 13 |
| Qwen3-30B-A3B-Instruct-2507-8bit-DWQ | 8-bit DWQ | 11.9 GB | 16 GB | 17 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.