4-bit Open on Hugging Face
Parameters
4.3B
Weights (est.)
≈ 3.0 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-4.0-3b-vision-nvfp4 --image photo.jpg --prompt "Describe this image"
Details
- Base model
- ibm-granite/granite-4.0-3b-vision (quantized)
- Vendor
- IBM · Granite
- Task
- Vision + text
- Quantisation
- 4-bit
- Licence
- apache-2.0
- Downloads
- 49 in the last 30 days · 860 all time · 0 likes
- Published
- 2026-04-01 · updated 2026-04-01 (7mo ago)
- Languages
- en
- Tags
- granite4_vision, custom_code
Other MLX builds of ibm-granite/granite-4.0-3b-vision
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| granite-4.0-3b-vision-4bit | 4-bit | ≈ 3.0 GB | 8 GB+ | 117 |
| granite-4.0-3b-vision-mxfp4 | MXFP4 | ≈ 3.0 GB | 8 GB+ | 52 |
| granite-4.0-3b-vision-nvfp4 (this one) | 4-bit | ≈ 3.0 GB | 8 GB+ | 49 |
| granite-4.0-3b-vision-5bit | 5-bit | ≈ 3.5 GB | 8 GB+ | 61 |
| granite-4.0-3b-vision-6bit | 6-bit | ≈ 4.0 GB | 8 GB+ | 37 |
| granite-4.0-3b-vision-8bit | 8-bit | ≈ 4.9 GB | 8 GB+ | 57 |
| granite-4.0-3b-vision-mxfp8 | MXFP8 | ≈ 4.9 GB | 8 GB+ | 32 |