Simplified-Reasoning · First MLX build 16 May 2026
SU-01
MLX
Text apache-2.0
A 31B parameter text generation model from Simplified-Reasoning, with 8 MLX builds, from MXFP4 to bf16.
- Parameters
- 31B From Hugging Face
- Quantisation
- MXFP4 SU-01-mxfp4
- Memory needed
- 22 GB Estimate, unified memory
- Download
- 17.2 GB Weights, estimated
- Downloads
- 19 This build, last 30 days
- Task
- Text generation Text
- MLX builds
- 8 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 22 GB at MXFP4
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/SU-01-mxfp4 --prompt "Hello"MLX builds
8 on mlx-community| SU-01-mxfp4 | MXFP4 | 17.2 GB | 22 GB | 19 |
|---|---|---|---|---|
| SU-01-4bit | 4-bit | 17.2 GB | 22 GB | 17 |
| SU-01-nvfp4 | 4-bit | 17.2 GB | 22 GB | 15 |
| SU-01-5bit | 5-bit | 21.0 GB | 26 GB | 17 |
| SU-01-6bit | 6-bit | 24.8 GB | 30 GB | 14 |
| SU-01-8bit | 8-bit | 32.4 GB | 39 GB | 17 |
| SU-01-mxfp8 | MXFP8 | 32.4 GB | 39 GB | 14 |
| SU-01-bf16 | bf16 | 61.1 GB | 71 GB | 15 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.