InvokeAI

Founded by Lincoln Stein in 2022

Generate and edit AI images locally on your own hardware

Cost

Free

Rating

★ People love it

Time to value

Quick Setup (< 1 hour)

You can use InvokeAI to generate AI images on your local machine without sending data to the cloud. It includes a layer-based canvas for painting and compositing, node-based workflow pipelines, and support for models like Flux, SDXL, and SD 1.5. You can manage LoRAs, ControlNets, and checkpoints through a built-in model manager. It runs on Windows, macOS, and Linux, and supports training custom models with its companion training UI.

What InvokeAI does

Generate images from text prompts using local AI modelsEdit and paint on AI-generated images using a layer-based canvasBuild custom image pipelines with a visual node editorManage and switch between multiple AI models and LoRAsTrain custom Textual Inversion or LoRA models on personal imagesApply ControlNet guidance using depth maps, edges, or pose dataRun batch image generation with prompt wildcardsUpscale generated images to higher resolutionsLayer-based canvas for painting, sketching, and compositing AI imagesNode-based workflow editor for building custom image generation pipelinesSupport for Flux, SDXL, and SD 1.5 models out of the boxBuilt-in model manager for checkpoints, LoRAs, ControlNets, and Textual InversionsRuns fully locally with no data sent to external serversBatch generation and prompt wildcards for rapid concept iterationCompanion training UI for Textual Inversion and LoRA model trainingApache 2.0 licensed and fully open-source