8-bit Open on Hugging Face
Parameters
31B
Weights (est.)
≈ 33.8 GB
Mac memory (est.)
48 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/gemma-4-31b-8bit --image photo.jpg --prompt "Describe this image"
Details
- Base model
- google/gemma-4-31b
- Vendor
- Google · Gemma
- Task
- Vision + text
- Quantisation
- 8-bit
- Licence
- apache-2.0
- Downloads
- 427 in the last 30 days · 49K all time · 24 likes
- Published
- 2026-04-02 · updated 2026-04-02 (7mo ago)
- Tags
- gemma4
Other MLX builds of google/gemma-4-31b
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| gemma-4-31b-4bit | 4-bit | ≈ 18.4 GB | 24 GB+ | 274 |
| gemma-4-31b-nvfp4 | 4-bit | ≈ 18.4 GB | 24 GB+ | 36 |
| gemma-4-31b-mxfp4 | MXFP4 | ≈ 18.4 GB | 24 GB+ | 33 |
| gemma-4-31b-5bit | 5-bit | ≈ 22.2 GB | 32 GB+ | 26 |
| gemma-4-31b-6bit | 6-bit | ≈ 26.1 GB | 32 GB+ | 62 |
| gemma-4-31b-8bit (this one) | 8-bit | ≈ 33.8 GB | 48 GB+ | 427 |
| gemma-4-31b-mxfp8 | MXFP8 | ≈ 33.8 GB | 48 GB+ | 34 |
| gemma-4-31b-bf16 | bf16 | ≈ 62.5 GB | 96 GB+ | 402 |