InvokeAI
InvokeAI is an open-source image generation tool under Apache-2.0 that runs as a locally hosted web server with a React interface. Its unified canvas combines generation, inpainting, outpainting and brush tools, and node workflows sit beside a simple form. The README lists 28 model families, including Stable Diffusion, SDXL, SD 3.5, Flux and Qwen Image. It installs through a launcher and serves as the base of several commercial products.
What does InvokeAI do?
InvokeAI's README describes "professional creative AI tools for visual media" that run as "a locally hosted web server & React UI." Its centre is the Unified Canvas, "a fully integrated canvas implementation with support for all core generation capabilities, in/out-painting, brush tools, and more." Node workflows combine "the power of node-based workflows with the ease of a UI."
The README lists 28 supported models, including SD 1.5, SD 2.0, SDXL, SD 3.5 Medium and Large, CogView 4, the Flux.1 family (Dev, Schnell, Kontext, Krea, Redux, Fill), Flux.2, Z-Image, Qwen Image and Qwen Image Edit. Three are marked API only: Nano Banana, GPT Image and Wan. It loads ckpt, diffusers and some GGUF models, keeps a gallery organised into boards, stores the prompt and settings in each image, and includes upscaling, a model manager and SAM segmentation. It says it is free to use under a commercially friendly licence.
Key facts
- Licence: Apache-2.0, with no added conditions; the README calls it a commercially friendly licence.
- Runs as a locally hosted web server with a React interface.
- Unified Canvas with generation, inpainting, outpainting and brush tools; node-based workflows.
- 28 model families listed, including SD 1.5, SDXL, SD 3.5, CogView 4, Flux.1 variants, Flux.2, Z-Image and Qwen Image; Nano Banana, GPT Image and Wan are marked API only.
- Model formats: ckpt, diffusers and some GGUF models.
- Gallery and boards; prompt and settings stored in each image; upscaling; SAM and SAM2 segmentation.
- Install through the Invoke Launcher; hardware requirements are not stated in the README.
- Latest tagged release when read: v6.14.1 on 2026-09-06; 28,237 stars on 2026-09-17.
What does it replace, and where does it fall short?
InvokeAI replaces Midjourney for someone who wants to work on an image, extending it, repainting parts, and combining generations on a canvas, with open models on their own machine and a licence that allows commercial use of the software. It is the pick for canvas-based editing on the open-source alternatives to Midjourney page.
Where it falls short: the README does not state hardware requirements or supported platforms, so check the launcher's documentation first. Three listed models run only through paid APIs. Video is limited to Wan through an API. For video and the widest model list, ComfyUI.
How does InvokeAI run?
Download the Invoke Launcher from GitHub and install from there; the README says to download and install on compatible hardware. It then runs a local web server you open in a browser. The README lists no hosted service; it says Invoke is the foundation for multiple commercial products and asks for GitHub Sponsors support.
Who is InvokeAI for?
An illustrator or designer who wants a canvas to compose, inpaint and extend images with open models. Someone who wants video or a node graph first should use ComfyUI.
What limits does the README state?
None stated in the README beyond support for "some" GGUF models and three models marked API only.
Questions people ask
Is InvokeAI open source?
Yes. InvokeAI is Apache-2.0 with no added conditions, and its README describes it as free to use under a commercially friendly licence. The image models you download carry their own licences, which you should check separately.
Can InvokeAI edit parts of an image?
Yes. Its Unified Canvas supports inpainting, which regenerates a masked area, outpainting, which extends an image beyond its edges, and brush tools, all in one canvas alongside generation. SAM segmentation helps select areas.
Which models does InvokeAI support?
The README lists 28, including SD 1.5, SD 2.0, SDXL, SD 3.5, CogView 4, the Flux.1 family, Flux.2, Z-Image, Qwen Image and Qwen Image Edit. Nano Banana, GPT Image and Wan are marked as API only, meaning they run through a paid service rather than locally.
How does InvokeAI compare with Midjourney?
InvokeAI gives you a canvas to generate and edit images with open models on your own computer, for free under Apache-2.0. Midjourney is a hosted service driven by prompts. The open-source alternatives to Midjourney page compares it with ComfyUI, SD.Next and SwarmUI.
What computer do I need for InvokeAI?
The README does not say; it tells you to install on compatible hardware and points to an FAQ for install problems. The launcher's documentation covers requirements, which depend on the models you run.
Sources
- InvokeAI README and Apache-2.0 LICENSE: github.com/invoke-ai/InvokeAI, read 2026-09-17.
Compared with the others
On the open-source alternatives to Midjourney page, InvokeAI is the pick for a canvas for inpainting and editing. Apache-2.0 with a unified canvas for generation, inpainting, outpainting and brushes, node workflows, and 28 model families.
Also on that page: ComfyUI for control, and the newest models, SD.Next for modest or unusual hardware, SwarmUI for a simple tab now, a graph later.
More image and video generation
The GPL-3.0 node-graph engine for image, video, audio and 3D generation that supports the newest open models, from Stable Diffusion and Flux to Wan and LTX video, on NVIDIA, AMD, Intel and Apple hardware, offline.
An Apache-2.0 all-in-one web interface for AI image and video generation, descended from AUTOMATIC1111's WebUI, with LoRA, ControlNet, quantisation that cuts video memory up to four times, and the widest hardware list here.
An MIT image and video generation web interface, formerly StableSwarmUI, with a simple Generate tab for beginners and a raw ComfyUI workflow tab for experts, that can spread work across several GPUs.
Added September 17, 2026. Every claim above comes from the project's README, LICENSE or model card, read on September 17, 2026, or from the GitHub API on the date shown in the panel. Found an error? Write to reveneau@licheo.com and it is fixed in the next weekly pass. Repository: github.com/invoke-ai/InvokeAI.
Running one of these inside your own environment, with your own data and your own security rules, is the kind of work Reveneau does. Read how a forward deployed engagement works.