IBM · First MLX build 28 Oct 2025
granite-4.0-h-350m
MLX
Text apache-2.0
A 340M parameter text generation model from IBM, with 6 MLX builds, from 3-bit to bf16.
- Parameters
- 340M From Hugging Face
- Quantisation
- 8-bit granite-4.0-h-350m-8bit
- Memory needed
- 3 GB Estimate, unified memory
- Download
- 362 MB Weights, estimated
- Downloads
- 734 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 3 GB at 8-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/granite-4.0-h-350m-8bit --prompt "Hello"MLX builds
6 on mlx-community| granite-4.0-h-350m-3bit | 3-bit | 149 MB | 3 GB | 28 |
|---|---|---|---|---|
| granite-4.0-h-350m-4bit | 4-bit | 192 MB | 3 GB | 217 |
| granite-4.0-h-350m-5bit | 5-bit | 234 MB | 3 GB | 66 |
| granite-4.0-h-350m-6bit | 6-bit | 277 MB | 3 GB | 20 |
| granite-4.0-h-350m-8bit | 8-bit | 362 MB | 3 GB | 734 |
| granite-4.0-h-350m-bf16 | bf16 | 681 MB | 3 GB | 22 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.