poolside · First MLX build 28 Apr 2026
Laguna-XS.2
MLX
Text apache-2.0
A 33B parameter text generation model from poolside, with 9 MLX builds, from 2-bit to bf16.
- Parameters
- 33B From Hugging Face
- Quantisation
- 4-bit Laguna-XS.2-4bit
- Memory needed
- 24 GB Estimate, unified memory
- Download
- 18.8 GB Weights, estimated
- Downloads
- 201 This build, last 30 days
- Task
- Text generation Text
- MLX builds
- 9 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 24 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/Laguna-XS.2-4bit --prompt "Hello"MLX builds
9 on mlx-community| Laguna-XS.2-2bit-M-DQ | 2-bit | 10.5 GB | 14 GB | 23 |
|---|---|---|---|---|
| Laguna-XS.2-4bit | 4-bit | 18.8 GB | 24 GB | 201 |
| Laguna-XS.2-nvfp4 | 4-bit | 18.8 GB | 24 GB | 43 |
| Laguna-XS.2-mxfp4 | MXFP4 | 18.8 GB | 24 GB | 31 |
| Laguna-XS.2-5bit | 5-bit | 23.0 GB | 28 GB | 27 |
| Laguna-XS.2-6bit | 6-bit | 27.2 GB | 33 GB | 38 |
| Laguna-XS.2-mxfp8 | MXFP8 | 35.5 GB | 42 GB | 50 |
| Laguna-XS.2-8bit | 8-bit | 35.5 GB | 42 GB | 48 |
| Laguna-XS.2-bf16 | bf16 | 66.9 GB | 77 GB | 26 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.