​

https://preview.redd.it/p44k1wnqtmeb1.png?width=1920&format=png&auto=webp&s=ae448e94c5913505d8e6fce39c2ac785c46f62c9

​

Hello, community! I'm happy to announce I have finally finished my ComfyUI SD Krita plugin. This is a plugin that allows users to run their favorite features from ComfyUI and at the same time, being able to work on a canvas.

Supports:

• Basic txt2img.
• Basic img2img.
• Inpainting (with auto-generated transparency masks).
• Simple upscale and upscaling with model (like Ultrasharp).
• Multicontrolnet with preprocessors. Annotator preview also available.
• Lora loading with Automatic1111 format.
• TI embeddings.
• (Advanced) Run custom ComfyUI workflow from Krita.
• (Advanced) Modify the default behaviour of each mode by injecting your own custom workflows.
• Import latest ComfyUI generations with one click.

This plugin offers a (somewhat) user friendly UI which can allow the user to make AI art in a rather simple way (for those that don't like the nodes), while also giving the option to use traditional digital art and edition tools to work on generated images. The plugin creates the necessary workflow under the hood to be run on an active ComfyUI instance. Generated images are then requested by the plugin and added to the canvas, and are also stored on your ComfyUI outputs directory. You are also given the possibility to work with your own custom workflows and inject them to the plugin to alter the behaviour of your generations.

JasonS09/comfy_sd_krita_plugin: Make AI art between canvas and nodes with Krita. (github.com)