LG AI Research · First MLX build 18 Mar 2025
EXAONE-Deep-32B
MLX
Text exaone
A 32B parameter text generation model from LG AI Research, with 6 MLX builds, from 3-bit to bf16.
- Parameters
- 32B From Hugging Face
- Quantisation
- 4-bit EXAONE-Deep-32B-4bit
- Memory needed
- 23 GB Estimate, unified memory
- Download
- 18.0 GB Weights, estimated
- Downloads
- 42 This build, last 30 days
- Task
- Text generation Text
- MLX builds
- 6 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 23 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/EXAONE-Deep-32B-4bit --prompt "Hello"MLX builds
6 on mlx-community| EXAONE-Deep-32B-3bit | 3-bit | 14.0 GB | 18 GB | 26 |
|---|---|---|---|---|
| EXAONE-Deep-32B-4bit | 4-bit | 18.0 GB | 23 GB | 42 |
| EXAONE-Deep-32B-6bit | 6-bit | 26.0 GB | 32 GB | 18 |
| EXAONE-Deep-32B-8bit | 8-bit | 34.0 GB | 41 GB | 22 |
| EXAONE-Deep-32B-mlx-8Bit | 8-bit | 34.0 GB | 41 GB | 15 |
| EXAONE-Deep-32B-bf16 | bf16 | 64.0 GB | 74 GB | 29 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.