4-bit Open on Hugging Face
Parameters
5.4B
Weights (est.)
≈ 4.5 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/gemma-3n-E2B-4bit --image photo.jpg --prompt "Describe this image"
Details
- Base model
- google/gemma-3n-E4B (quantized)
- Vendor
- Google · Gemma
- Task
- Vision + text
- Quantisation
- 4-bit
- Licence
- gemma
- Downloads
- 68 in the last 30 days · 1.8K all time · 4 likes
- Published
- 2025-07-06 · updated 2026-01-21 (9mo ago)
- Tags
- gemma3n, automatic-speech-recognition, automatic-speech-translation, audio-text-to-text, video-text-to-text
Other MLX builds of google/gemma-3n-E4B
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| gemma-3n-E2B-3bit | 3-bit | ≈ 3.9 GB | 8 GB+ | 25 |
| gemma-3n-E2B-4bit (this one) | 4-bit | ≈ 4.5 GB | 8 GB+ | 68 |
| gemma-3n-E4B-it-3bit | 3-bit | ≈ 5.0 GB | 8 GB+ | 31 |
| gemma-3n-E2B-5bit | 5-bit | ≈ 5.0 GB | 8 GB+ | 33 |
| gemma-3n-E2B-bf16 | bf16 | ≈ 10.9 GB | 16 GB+ | 70 |
| gemma-3n-E4B-it-bf16 | bf16 | ≈ 15.7 GB | 24 GB+ | 161 |
| gemma-3n-E4B-it-4bit | 4-bit | — | — | 425 |
| gemma-3n-E4B-it-5bit | 5-bit | — | — | 55 |
| gemma-3n-E4B-it-8bit | 8-bit | — | — | 50 |
| gemma-3n-E2B-8bit | 8-bit | — | — | 40 |
| gemma-3n-E4B-it-6bit | 6-bit | — | — | 32 |
| gemma-3n-E2B-6bit | 6-bit | — | — | 23 |