MLX Community · First MLX build 5 Jun 2026
gemma-4-26B-A4B-it-qat
MLX
TextVision
A 26B parameter vision and text model, with 8 MLX builds, from 4-bit to bf16.
- Parameters
- 26B From Hugging Face
- Quantisation
- 4-bit gemma-4-26B-A4B-it-qat-4bit
- Memory needed
- 20 GB Estimate, unified memory
- Download
- 15.3 GB Weights, estimated
- Downloads
- 5.9K This build, last 30 days
- Task
- Vision and text Text, Vision
- MLX builds
- 8 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 20 GB at 4-bit
8 GB
16 GB
24 GB
32 GB
64 GB
128 GB
Fits with headroomTight, close other appsWon’t fit
Estimated from the size of the weights, plus a fifth for context and working memory and 2 GB for macOS and your apps. Long contexts need more.
Run it
Terminal · mlx-vlm$ pip install -U mlx-vlm
$ python -m mlx_vlm.generate --model mlx-community/gemma-4-26B-A4B-it-qat-4bit --image photo.jpg --prompt "Describe this image"MLX builds
8 on mlx-community| gemma-4-26B-A4B-it-qat-4bit | 4-bit | 15.3 GB | 20 GB | 5.9K |
|---|---|---|---|---|
| gemma-4-26B-A4B-it-qat-mxfp4 | MXFP4 | 15.3 GB | 20 GB | 168 |
| gemma-4-26B-A4B-it-qat-nvfp4 | 4-bit | 15.3 GB | 20 GB | 153 |
| gemma-4-26B-A4B-it-qat-5bit | 5-bit | 18.5 GB | 23 GB | 65 |
| gemma-4-26B-A4B-it-qat-6bit | 6-bit | 21.6 GB | 27 GB | 183 |
| gemma-4-26B-A4B-it-qat-8bit | 8-bit | 28.0 GB | 34 GB | 548 |
| gemma-4-26B-A4B-it-qat-mxfp8 | MXFP8 | 28.0 GB | 34 GB | 81 |
| gemma-4-26B-A4B-it-qat-bf16 | bf16 | 51.6 GB | 60 GB | 112 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.