Generate concept art and illustrations from text descriptions for games, animation, or design projects.
Edit and refine existing images by inpainting details or extending scenes beyond original borders.
Build custom multi-step image generation workflows by connecting AI components visually without coding.
Organize and manage a library of AI-generated images with built-in gallery tools while keeping all work private.
Requires downloading large model files (multi-GB), GPU/CUDA setup, and coordinating Python backend with React frontend.
InvokeAI is a locally-run, browser-based creative tool for generating images and visual media using AI image generation models (specifically Stable Diffusion and related models, which use a technique called "latent diffusion" to create images from text descriptions). The tool runs entirely on your own computer, you download it, install it, and it starts a local web server that you access in your browser. No data is sent to external services. It features a polished web interface with several core capabilities: text-to-image generation (describe an image and the AI creates it), inpainting (edit or fill in parts of an existing image), and an outpainting canvas (extend images beyond their original borders). There's also a node-based workflow editor, meaning you can build custom multi-step image generation pipelines by connecting components visually, like a flow diagram. The tool supports a wide range of AI model types, including Stable Diffusion 1.5, SDXL, SD 3.5, and Flux models. A built-in model manager lets you download and switch between models. It also includes gallery management to organize your generated images, and supports advanced techniques like upscaling (increasing image resolution), embeddings (style or concept add-ons), and object segmentation models. You'd use InvokeAI if you're an artist, designer, or creative professional who wants a capable, privacy-friendly image generation studio running locally. It's free to use under a commercially-friendly license and is written primarily in TypeScript and Python.
Generated 2026-05-18 · Model: sonnet-4-6 · Verify against the repo before relying on details.