Sign In
home
models
images
videos
posts
articles
bounties
tools
challenges
events
shop

Krita AI-DIffusion - A plugin with Stable Diffusion inside Krita

MisterMR's Avatar
MisterMR
Gold Writer Badge
Feb 11, 2024
tool guide
kritaai diffusion

Objective

Testing the Krita AI-Diffusion component to edit images in Krita.

Preface

I never use this component before, I decide to write the article while I was doing my first installation.

Software

Installation

Krita

You can download Krita from https://krita.org/en/download/krita-desktop/

For the installation check the manuals:

https://docs.krita.org/en/user_manual/getting_started/installation.html

I use a Krita for drawing (to avoid buying commercial software because it is expensive) and also because reads .PSD files and export in several formats.

When I develop games you can put .PSD files in unity folder and change them with Krita.

Krita has layers and are very good with several options.

Also the brushes are very cool, you have so many and you can personalize them.

If you have a Pen Tablet or Pen Display Krita works very well with pen pressure.

Plug-in krita-ai-diffusion

On the page https://github.com/Acly/krita-ai-diffusion you can reAd the installation ( I added here with some pictures)

Installation

  1. If you haven't yet, go and install Krita! Required version: 5.2.0 or newer

  2. Download the plugin. Unpack the archive into your pykrita folder.

    • Windows: Usually C:\Users\<user>\AppData\Roaming\krita\pykrita

    • Linux: Usually ~/.local/share/krita/pykrita

    • MacOS: Usually ~/Library/Application Support/krita/pykrita

    • Check Krita's official documentation if you have trouble locating it.

  3. Enable the plugin in Krita (Settings ‣ Configure Krita ‣ Python Plugins Manager) and restart.

  4. Create a new document or open an existing image.

  5. To show the plugin docker: Settings ‣ Dockers ‣ AI Image Generation.

  6. In the plugin docker, click "Configure" to start server installation. Requires 10+ GB free disk space.

    You can decide if to use a local version of ConfyUI or install a new version for this software

    I decide to install the new version.

The default settings install SD 1.5 and as a checkpoint Realistic Vision and DreamShaper.

Because it is a ComfyUI installation I think later I can add some of the checkpoint that I like.

I decided to install in a new path , it can take a while to download 10GB

View logs

If you press "View log files" to check what the system is installing..

Installation completed

When the download/installation it is completed you will see the server running

and the components all in green.

Using the component

Create an empty document from File >> New

I selected 768x512 but you can select the size you want.. I think that with SD 1.5 this is one of the better size for not having trouble with image generation

Now we can write in the prompt

And you can select if to use Cinematic Photo or Digital Artwork.

There are also 3 options : Generate, Upscale and Live.

I try with the Generate option and s simple prompt: A beautiful tropical background in a sunny day with a lot of colors

I get 5 different options.

when I Select one

If you user the selection tools you can

You have more option to work with

I did a selection and after I wen to menù Select >> Invert Selection

A this point I wrote a new prompt and I run 1girl blonde with red bikini full body

Quality is not the best (yet)

I now impressed with the image result, but with the power of having image generation inside my favorite tool.

I will try the component in the next days, I think will be great for image generation and changes for my games.

Thanks for reading this tutorial.

8

Comments