invoke-ai/InvokeAI

Invoke is a leading creative engine for Stable Diffusion models, empowering professionals, artists, and enthusiasts to generate and create visual media using the latest AI-driven technologies. The solution offers an industry leading WebUI, and serves as the foundation for multiple commercial products

Summary

Invoke is an open-source, self-hosted creative engine for image generation models including Stable Diffusion, SDXL, Flux.1, Flux.2, CogView 4, and others. It runs a local web server with a React-based UI and a node-based workflow system, and the project is released under the Apache-2.0 license.

What it is useful for

• Running text-to-image, image-to-image, inpainting, and outpainting on your own machine without depending on a hosted service.
• Building and sharing customizable generation pipelines through a node-based workflow editor, which the README describes as combining "the power of node-based workflows with the ease of a UI."
• Iterating on imagery in the Unified Canvas, which supports in/out-painting, brush tools, and overlay on existing sketches, photography, or renders.
• Managing a local model library across SD 1.5, SD 2.0, SDXL, SD 3.5 (Medium/Large), Flux.1 family (Dev, Schnell, Kontext, Krea, Redux, Fill), Flux.2 Dev, Flux.2 Klein (4B and 9B), Z-Image (Turbo/Base), Krea 2, Qwen Image, Ideogram 4, ERNIE-Image, and Anima, in ckpt, diffusers, and some gguf formats.
• Selecting regions with SAM / SAM2, upscaling generated images, and managing embeddings from inside the app.
• Storing outputs in a local gallery that preserves prompt and settings metadata for later recall or remixing.
• Using it as a foundation layer; the README notes it "serves as the foundation for multiple commercial products."

How engineers can use it

1. Download the Launcher from the linked releases page and follow the installation docs at invoke.ai/start-here/installation/ to install on Windows, macOS, or Linux.
2. Launch the app and open the WebUI in a browser; use the Model Manager to fetch a supported checkpoint, then run a first generation.
3. Open the Workflows editor to chain nodes into a reusable pipeline, or use the Unified Canvas to inpaint or outpaint an existing image.
4. The troubleshooting FAQ at invoke.ai/troubleshooting/faq/ covers common install issues, and the Discord linked in the README is the documented escalation path.

Documented limitations and caveats:

• Several listed models (Nano Banana, GPT Image, Wan) are marked "API Only" in the README, so they are not runnable locally through this engine.
• The README only states "compatible hardware"; specific GPU/driver prerequisites are not enumerated in the supplied source, so consult the installation docs before committing.
• Setup beyond "download the Launcher" is not spelled out in the README itself; the docs site is the authoritative installation reference.