One open-source workspace
for AI images and videos
Open Generative AI brings image generation, video generation, lip sync, cinematic controls and more than 200 AI models into one interface. An open alternative to tools like Higgsfield, without 99 dollars every month.
- The app is free and open source, MIT licensed
- Cloud generations may consume paid API credits
- No mandatory monthly subscription for the app
A full studio in one app
Image generation
Text-to-image and image-to-image across models like Flux, Midjourney and Seedream.
Text-to-video
Describe the scene and get a video through models like Kling, Sora and Veo.
Image-to-video
Animate any still image into a video shot with camera motion you describe.
Lip sync
A dedicated studio with nine models: a portrait plus audio, or an existing video plus new audio.
Cinema Studio
Photorealistic cinematic shots with pro controls: lens, focal length and aperture.
More than 200 models
All the big models in one place, switch between them from the same interface.
Get it on your machine
One-click installers, no Node.js and no terminal. Grab your copy from the official releases page.
- 1Download the Setup .exe from the releases page
- 2SmartScreen may warn you because the installer is not code-signed: click More info, then Run anyway
- 3The app installs silently with a Start Menu shortcut
Five steps and you are in
- 1Create or obtain a MuAPI key from muapi.ai
- 2Open the Open Generative AI app
- 3Add the API key in settings
- 4Pick the model you like
- 5Start generating
The app is open source, but cloud image and video generations can consume API credits.
Your first generation
Open the image studio, paste this prompt, and pick an image model like Flux:
A cinematic product shot of a futuristic AI assistant floating above a dark creator desk, premium commercial lighting, realistic materials, vertical 9:16
Then take the result into image-to-video and animate it with this description:
Slow cinematic camera push-in, subtle orbit, realistic reflections, stable subject, no morphing
Connect it with Claude Code
The same media ecosystem has a companion called Generative Media Skills: a skills library that lets Claude Code generate and edit media straight from the terminal, with the same models.
- Open Generative AI is the visual interface
- Claude Code uses the companion Generative Media Skills
- Claude improves prompts, picks workflows, and generates media from the terminal
- Claude does not directly control the desktop interface
Three commands and you are set:
$ npm install -g muapi-cli $ muapi auth configure --api-key "YOUR_MUAPI_KEY" $ npx skills add SamurAIGPT/Generative-Media-Skills --all -a claude-code
Start creating without another monthly subscription
Open Generative AI is open source, MIT licensed. Cloud model usage may require paid API credits. GitHub