VibeusKrita AI Diffusion
An open-source Krita plugin for inpainting, outpainting, text-guided edits and AI-assisted painting, with local or cloud diffusion workflows.
Krita AI Diffusion is an open-source plugin that brings generative image editing into Krita. Work with selections and layers instead of leaving the canvas for a separate image-generation app: use generative fill to add or remove content, extend an image, describe different regions, and guide results with sketches, line art, depth, pose and reference images. The plugin also supports live painting, text-guided edits and upscaling, with a queue and history for jobs and prompts.
ComfyUI is the diffusion backend. The README describes an automatic local setup as well as connecting to a compatible existing or remote ComfyUI server. Listed model families include Flux 2, Z-Image and Stable Diffusion 1.5/XL; ControlNet and IP-Adapter provide additional control and reference-image workflows when the required models and extensions are configured.
Krita 5.2 or newer is required. Windows, Linux and macOS are supported. For local generation, the project recommends a GPU with at least 6 GB of VRAM; CPU mode is supported but very slow, and limited memory may cause generation to fail. Cloud generation is another way to get started without a powerful local machine. The plugin is distributed from the project’s releases under GPL-3.0.
