---
license: other
license_name: ltx-video-license
license_link: https://huggingface.co/Lightricks/LTX-2/blob/main/LICENSE.md
library_name: mlx
tags:
- video-generation
- mlx
- apple-silicon
- ltx-2
- dev
- quantized
---

# LTX-2 19B Dev (4-bit) - MLX

This is a 4-bit quantized version of the [LTX-2 19B Dev](https://huggingface.co/Lightricks/LTX-2) model, optimized for Apple Silicon using MLX.

## Model Description

LTX-2 is a state-of-the-art video generation model from Lightricks. This version has been quantized to 4-bit precision for efficient inference on Apple Silicon devices with MLX.

### Key Features

- **Pipeline**: Dev (full control with CFG scale)
- **Quantization**: 4-bit precision
- **Framework**: MLX (Apple Silicon optimized)
- **Memory**: ~19GB VRAM required

## Usage

### Installation

```bash
pip install git+https://github.com/CharafChnioune/mlx-video.git
```

### Command Line

```bash
# Basic generation
mlx-video --prompt "A beautiful sunset over the ocean" \
    --model-repo AITRADER/ltx2-dev-4bit-mlx \
    --pipeline dev \
    --height 512 --width 512 \
    --num-frames 33

# Dev pipeline with CFG
mlx-video --prompt 'A cat playing with yarn' \\
    --model-repo AITRADER/ltx2-dev-4bit-mlx \\
    --pipeline dev \\
    --steps 40 --cfg-scale 4.0
```

### Python API

```python
from mlx_video import generate_video

video = generate_video(
    prompt="A beautiful sunset over the ocean",
    model_repo="AITRADER/ltx2-dev-4bit-mlx",
    pipeline="dev",
    height=512,
    width=512,
    num_frames=33,
)
```

## Model Files

- `ltx-2-19b-dev-mlx.safetensors` - Main model weights (4-bit quantized)
- `quantization.json` - Quantization configuration
- `config.json` - Model configuration
- `layer_report.json` - Layer information

## Performance

| Resolution | Frames | Steps |
|------------|--------|------|
| 512x512 | 33 | ~40 steps |
| 768x512 | 33 | ~40 steps |

## License

This model is released under the [LTX Video License](https://huggingface.co/Lightricks/LTX-2/blob/main/LICENSE.md).

## Acknowledgements

- [Lightricks](https://lightricks.com/) for the original LTX-2 model
- [MLX](https://github.com/ml-explore/mlx) team at Apple for the framework
- [mlx-video](https://github.com/CharafChnioune/mlx-video) for the MLX conversion