InvokeAI

InvokeAI is a free, open-source creative engine for AI image generation. It is built for creatives and is described as self-hosted, fully customizable, and licensed under Apache 2.0.

Its documentation highlights a set of features aimed at production-oriented image workflows. Invoke includes a layer-based canvas for drawing, painting, sketching, and editing, with independently manipulable layers for composing scenes. It also offers node-based workflows for building complex, reproducible pipelines through a visual graph backend, along with custom UI parameters that can be exposed for sharing and updating values. The model manager supports checkpoints, LoRAs, Textual Inversions, and ControlNets, and the software provides out-of-the-box support for Flux, SDXL, SD 1.5, and other foundational models.

The application is described as completely local and self-hosted, running on the user’s own hardware so that data, prompts, and creations remain under the user’s control. It is available for Windows, macOS, and Linux, with the Invoke Launcher presented as the fastest way to get started on those platforms. For advanced setups, the documentation mentions Docker and manual Python installation, and it also lists hosted GPU options on AI Badgr, RunPod, and Railway. The Windows download requires Windows 10 or later and an NVIDIA or AMD GPU; macOS requires Apple Silicon; and Linux is compatible with most distributions while also requiring an NVIDIA or AMD GPU.

Invoke also includes features for iteration and model training. The documentation says batch generation, prompt wildcards, and high-resolution upscaling are available in the app, and Invoke Training provides a dedicated UI for Textual Inversion and LoRA training. The project is actively developed by an open-source community, with support, contribution guidance, and community channels mentioned on the site.