InvokeAI is an open-source creative engine for AI image generation that runs as a locally hosted web server with a React interface. It supports a long list of models including SD 1.5, SDXL, SD 3.5, Flux.1 and Flux.2 families, Qwen Image, Z-Image and others, plus API-only options. The Unified Canvas brings generation, inpainting, outpainting and brush tools into one workspace, and a node-based workflow editor lets users build reusable pipelines. It is free to use under the Apache-2.0 license, which is described as commercially friendly.
Digital artists and illustrators use it to refine and iterate on images, mixing AI output with sketches, photography and renders. Studios and technical directors build shareable workflows for specific production needs. Hobbyists and developers who want control over their models, embeddings and data appreciate that everything runs on their own hardware after installation with the Launcher. Contributors can join the open-source project on GitHub, and the community answers questions on Discord, which helps newcomers get through installation.
Typical workflows include generating concept art with a Flux or SDXL model, then moving to the canvas to inpaint a face, extend the frame with outpainting or upscale the result. A game artist can save a node workflow that applies a consistent style to many assets, while a photographer can use segmentation models to select objects before editing them. The gallery stores boards of images with rich metadata, so prompts and settings from a past image can be recalled and remixed later. Teams building commercial products also use it as a foundation, since the project serves as the base for several commercial offerings.
