MXFP4 Open on Hugging Face
Parameters
3.1B
Weights (est.)
≈ 2.3 GB
Mac memory (est.)
8 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-vlm on an Apple silicon Mac:
pip install mlx-vlmpython -m mlx_vlm.generate --model mlx-community/granite-vision-3.2-2b-mxfp4 --image photo.jpg --prompt "Describe this image"
Details
- Base model
- ibm-granite/granite-vision-3.2-2b (quantized)
- Vendor
- IBM · Granite
- Task
- Vision + text
- Quantisation
- MXFP4
- Licence
- apache-2.0
- Downloads
- 47 in the last 30 days · 238 all time · 0 likes
- Published
- 2026-04-01 · updated 2026-04-01 (7mo ago)
- Languages
- en
- Tags
- granite_vision
Other MLX builds of ibm-granite/granite-vision-3.2-2b
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| granite-vision-3.2-2b-4bit | 4-bit | ≈ 2.3 GB | 8 GB+ | 75 |
| granite-vision-3.2-2b-nvfp4 | 4-bit | ≈ 2.3 GB | 8 GB+ | 63 |
| granite-vision-3.2-2b-mxfp4 (this one) | MXFP4 | ≈ 2.3 GB | 8 GB+ | 47 |
| granite-vision-3.2-2b-5bit | 5-bit | ≈ 2.7 GB | 8 GB+ | 35 |
| granite-vision-3.2-2b-6bit | 6-bit | ≈ 3.0 GB | 8 GB+ | 38 |
| granite-vision-3.2-2b-8bit | 8-bit | ≈ 3.7 GB | 8 GB+ | 43 |
| granite-vision-3.2-2b-mxfp8 | MXFP8 | ≈ 3.7 GB | 8 GB+ | 35 |
| granite-vision-3.2-2b-bf16 | bf16 | ≈ 6.2 GB | 16 GB+ | 45 |