MXFP4 Open on Hugging Face
Parameters
29B
Weights (est.)
≈ 16.2 GB
Mac memory (est.)
24 GB+
Size is estimated from the parameter count and quantisation. Memory allows a fifth extra for context plus 2 GB for the system; long contexts need more.
Run it
With mlx-lm on an Apple silicon Mac:
pip install mlx-lmmlx_lm.generate --model mlx-community/granite-4.1-30b-mxfp4 --prompt "Hello"
Details
- Base model
- ibm-granite/granite-4.1-30b (quantized)
- Vendor
- IBM · Granite
- Task
- Text generation
- Quantisation
- MXFP4
- Licence
- apache-2.0
- Downloads
- 39 in the last 30 days · 1.9K all time · 2 likes
- Published
- 2026-04-30 · updated 2026-04-30 (6mo ago)
- Tags
- granite, language, granite-4.0
Other MLX builds of ibm-granite/granite-4.1-30b
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| granite-4.1-30b-4bit | 4-bit | ≈ 16.2 GB | 24 GB+ | 675 |
| granite-4.1-30b-nvfp4 | 4-bit | ≈ 16.2 GB | 24 GB+ | 47 |
| granite-4.1-30b-mxfp4 (this one) | MXFP4 | ≈ 16.2 GB | 24 GB+ | 39 |
| granite-4.1-30b-5bit | 5-bit | ≈ 19.8 GB | 32 GB+ | 23 |
| granite-4.1-30b-6bit | 6-bit | ≈ 23.5 GB | 32 GB+ | 24 |
| granite-4.1-30b-mxfp8 | MXFP8 | ≈ 30.7 GB | 48 GB+ | 69 |
| granite-4.1-30b-8bit | 8-bit | ≈ 30.7 GB | 48 GB+ | 48 |
| granite-4.1-30b-bf16 | bf16 | ≈ 57.7 GB | 96 GB+ | 38 |