Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
GIMP plugin for AUTOMATIC1111's Stable Diffusion WebUI
| Date | Stars |
|---|---|
| 2026-07-31 | 403 |
| 2026-08-06 | 403 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
<a name="readme-top"></a>
<div>
<h1 align="center">Stable Boy</h1>
<p align="center">
A GIMP plugin for AUTOMATIC1111's Stable Diffusion WebUI
</p>
</div>
## About
[AUTOMATIC1111's Stable Diffusion WebUI](https://github.com/AUTOMATIC1111/stable-diffusion-webui) is one of the most powerful tools in the generative AI space. Stable Boy puts that power into GIMP 2.10 by calling into A1111 WebUI's API.
Here's a short demo video (of v0.1 ... haven't found time to make a new one):
[](https://youtu.be/YMVog30OcTI)
<p align="right">(<a href="#readme-top">back to top</a>)</p>
## Installation
1. Clone the repo:
```sh
git clone https://github.com/thndrbrrr/gimp-stable-boy.git
```
1. Start GIMP, open Preferences and add the absolute path to _sub-folder_ (!) `src` to GIMP's plugin search path:
```
/home/...and-so-on.../gimp-stable-boy/src # MacOS, Linux
C:/Users/...and-so-on.../gimp-stable-boy/src # Windows
```
1. Restart GIMP.
1. Start A1111 WebUI with argument `--api`. It doesn't matter whether it's running on your local machine or somewhere in the cloud.
1. Open Stable Boy's preferences (top menu > Stable Boy > Preferences) and copy-paste A1111 WebUI's URL (something like `https://abcdef123.gradio.com`) into the `API URL` field.
<p align="right">(<a href="#readme-top">back to top</a>)</p>
## Usage
Stable Boy can be found in the `Stable Boy` top menu, with the following options:
- Preferences
- Text to Image
- Image to Image
- Inpainting
- Upscale
- Scripts
- X/Y plot
> ℹ️ Stable Boy's menu items will be grayed out if no image is open.
### Model selection
Stable Boy currently doesn't have an option to choose the model. You'll have to do that in WebUI. Loading an inpainting model is recommended.
### Image size
The minimum resolution is 512x512. Ideally choose dimensions that are multiples of 256.
### Generate images or layers
Stable Boy can open results coming from Stable Diffusion as new images, or they can be added as new layers to the current image. The `Results as` option dropdown lets you control this.
Using layers is very powerful, especially when inpainting large images.
### Inpainting
Add a layer named `Inpainting Mask` to the image and make it the top layer. Use a paintbrush and paint the region you want to inpaint with black on that mask layer. Stable Boy will automatically determine the area of the image to process (multiples of 256 and at least 512x512). When results are added as layers, the inpainting mask is applied to those layers so that they really only contain the masked part.
In the GIF below you can see how multiple variations of inpainting are added as layers, which can then be compared easily:
<!--  -->
Stable Boy inserts any generated layers below the inpainting mask layer and then hides the mask layer. The mask itself remains as is.
> 👉 Select the mask layer and hit `DEL` to clear the mask.
### X/Y plotting
X/Y plotting is an essential step in generative art workflows: since changes in parameters are not immediate and multiple parameters often work in conjunction, the best way to go about it is to compare the results of different settings in an X/Y plot. When using Image to Image for example, an X/Y plot of multiple CFG scales and denoising strengths makes it very easy to determine the right combination:
The result of X/Y plot is a grid as well as the individual images. Use the `Grid only` option to control whether the individual images should be included. As can be seen in the GIF image above, grids are always opened as separate images, whereas individual result images are added to the existing image as nested layer groups. Each layer is nameExcerpt of 7,653 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:15d166ef08b66b05, desc:stable diffusion