Alibaba Qwen · First MLX build 30 Jul 2025
Qwen3-30B-A3B-Thinking-2507
MLX
Text apache-2.0
A 31B parameter text generation model from Alibaba Qwen, with 2 MLX builds, from 4-bit to bf16.
- Parameters
- 31B From Hugging Face
- Quantisation
- 4-bit Qwen3-30B-A3B-Thinking-2507-4bit
- Memory needed
- 22 GB Estimate, unified memory
- Download
- 17.2 GB Weights, estimated
- Downloads
- 199 This build, last 30 days
- Task
- Text generation Text
- MLX builds
- 2 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 22 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 · mlx-lm$ pip install -U mlx-lm
$ mlx_lm.generate --model mlx-community/Qwen3-30B-A3B-Thinking-2507-4bit --prompt "Hello"MLX builds
2 on mlx-community| Qwen3-30B-A3B-Thinking-2507-4bit | 4-bit | 17.2 GB | 22 GB | 199 |
|---|---|---|---|---|
| Qwen3-30B-A3B-Thinking-2507-bf16 | bf16 | 61.1 GB | 71 GB | 54 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.