IBM · First MLX build 3 May 2026
granite-4.1-8b
MLX
Text apache-2.0
A 8.4B parameter text generation model from IBM, with 8 MLX builds, from 4-bit to bf16.
- Parameters
- 8.4B From Hugging Face
- Quantisation
- 4-bit granite-4.1-8b-4bit
- Memory needed
- 8 GB Estimate, unified memory
- Download
- 4.7 GB Weights, estimated
- Downloads
- 826 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 8 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/granite-4.1-8b-4bit --prompt "Hello"MLX builds
8 on mlx-community| granite-4.1-8b-4bit | 4-bit | 4.7 GB | 8 GB | 826 |
|---|---|---|---|---|
| granite-4.1-8b-nvfp4 | 4-bit | 4.7 GB | 8 GB | 46 |
| granite-4.1-8b-mxfp4 | MXFP4 | 4.7 GB | 8 GB | 35 |
| granite-4.1-8b-5bit | 5-bit | 5.8 GB | 9 GB | 31 |
| granite-4.1-8b-6bit | 6-bit | 6.8 GB | 10 GB | 41 |
| granite-4.1-8b-8bit | 8-bit | 8.9 GB | 12 GB | 517 |
| granite-4.1-8b-mxfp8 | MXFP8 | 8.9 GB | 12 GB | 35 |
| granite-4.1-8b-bf16 | bf16 | 16.8 GB | 21 GB | 40 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.