Metadata-Version: 2.4
Name: gemini-unwatermark
Version: 0.1.1
Summary: Remove watermarks from Gemini-generated images using LaMa inpainting
Project-URL: Homepage, https://github.com/jaylann/gemini-unwatermark
Project-URL: Repository, https://github.com/jaylann/gemini-unwatermark
Author: Justin Lanfermann
License-Expression: Apache-2.0
License-File: LICENSE
Keywords: ai,gemini,image-processing,inpainting,lama,watermark
Classifier: Development Status :: 4 - Beta
Classifier: Environment :: Console
Classifier: Intended Audience :: End Users/Desktop
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Topic :: Multimedia :: Graphics
Requires-Python: <3.14,>=3.10
Requires-Dist: httpx>=0.25.0
Requires-Dist: numpy>=1.24.0
Requires-Dist: onnxruntime>=1.16.0
Requires-Dist: pillow>=10.0.0
Requires-Dist: rich>=13.0.0
Requires-Dist: typer>=0.9.0
Provides-Extra: gpu
Requires-Dist: onnxruntime-gpu>=1.16.0; extra == 'gpu'
Description-Content-Type: text/markdown

# gemini-unwatermark

Fast CLI to remove watermarks from Google Gemini-generated images using AI inpainting.

## Installation

```bash
pip install gemini-unwatermark
```

Or run directly without installing:
```bash
uvx gemini-unwatermark image.png
```

For GPU acceleration (CUDA):
```bash
pip install "gemini-unwatermark[gpu]"
```

**Requirements:** Python 3.10-3.13

## Usage

```bash
# Single image
gemini-unwatermark image.png

# Custom output
gemini-unwatermark image.png -o clean.png

# Batch processing
gemini-unwatermark *.png -d ./cleaned/

# Parallel (8 workers)
gemini-unwatermark photos/*.jpg -d ./output -w 8
```

## Options

| Option | Description |
|--------|-------------|
| `-o, --output` | Output path (single image only) |
| `-d, --output-dir` | Output directory for batch |
| `-f, --force` | Overwrite existing files |
| `-q, --quiet` | Suppress progress output |
| `-w, --workers` | Parallel workers (1-8, default: 4) |
| `--model-path` | Custom ONNX model path |

## How it works

Uses [LaMa](https://github.com/advimman/lama) (Large Mask Inpainting) to intelligently fill the watermark region in the bottom-right corner of Gemini images.

The model (~200MB) auto-downloads on first use to `~/.cache/gemini-unwatermark/`.

## Disclaimer

This tool only removes the **visible** watermark. It does **not** remove [SynthID](https://deepmind.google/technologies/synthid/), Google's invisible watermarking technology embedded in the image data.

## Credits

- Original Chrome extension: [dinoBOLT/Gemini-Watermark-Remover](https://github.com/dinoBOLT/Gemini-Watermark-Remover)
- LaMa ONNX model: [Carve/LaMa-ONNX](https://huggingface.co/Carve/LaMa-ONNX)

## License

Apache-2.0
