explaingit

anil-matcha/open-generative-ai

13,242JavaScriptAudience · designerComplexity · 2/5LicenseSetup · easy

TLDR

A free, open-source desktop and browser app for generating images and videos with AI, connects to 200+ models including Flux, Midjourney, and Sora, with no content filters.

Mindmap

mindmap
  root((repo))
    What It Does
      Image generation
      Video generation
      Lip sync animation
    Studios
      Image Studio
      Video Studio
      Cinema Studio
      Lip Sync Studio
    Tech Stack
      JavaScript
      Electron
      Muapi.ai models
    How to Use
      Hosted browser app
      Desktop installer
      Self-hosted option
Click or tap to explore — scroll the page freely

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

Things people build with this

USE CASE 1

Generate images from text prompts using Flux, Midjourney, or 200+ other models through one interface.

USE CASE 2

Animate a portrait photo or sync facial lip movements to an audio clip using the Lip Sync studio.

USE CASE 3

Self-host the tool on your own machine and add custom AI models or modify the interface.

USE CASE 4

Turn a text description into a video clip using models like Kling, Sora, or Veo without a paid subscription.

Tech stack

JavaScriptElectron

Getting it running

Difficulty · easy Time to first run · 5min

Desktop app is unsigned, macOS and Windows show a security warning that requires a manual override to open.

MIT, use freely for any purpose, including commercial, as long as you keep the copyright notice.

In plain English

Open Generative AI is a free, open-source program for creating images and videos with artificial intelligence. The README presents it as an alternative to commercial AI media platforms, and it groups its features into four studios: Image, Video, Cinema, and Lip Sync. The last of these can animate a portrait or match a face's lip movements to a piece of audio. Rather than running the AI models itself by default, the tool connects to a large catalog of more than 200 existing models, named ones include Flux, Midjourney, Kling, Sora, and Veo, through a service called Muapi.ai. It supports common tasks such as turning text into an image, changing an existing image, turning text or an image into a video, and feeding several reference images into models that accept them. The desktop version can also run some image models locally on your own computer using two optional engines, one bundled and one that you host separately on a machine with a suitable graphics card. There are several ways to use it. The README links to a hosted version you can try in a browser with a free account, and offers one-click desktop installers for macOS, Windows, and Linux. Because the desktop app is not signed by Apple or Microsoft, the README walks through the security warnings each operating system shows and how to get past them, including specific steps for newer Ubuntu versions. The project states plainly that it applies no content filters or prompt restrictions, which the author frames as full creative freedom. Readers should be aware this also means the usual safety guardrails found on commercial services are absent. The code is released under the MIT license, can be self-hosted, and is described as extensible so that users can add their own models or change the interface. The README also points to related projects from the same author and to community channels on Reddit and Discord.

Copy-paste prompts

Prompt 1
I'm using open-generative-ai. Help me write a text prompt that generates a realistic product photo on a white background using Flux.
Prompt 2
Using open-generative-ai on macOS, I get a security warning when opening the app. Walk me through the exact steps to open it anyway.
Prompt 3
I want to add a custom AI model to open-generative-ai. Show me where in the codebase to register it and what format it expects.
Prompt 4
Help me set up the local image generation engine in open-generative-ai on a Windows machine with an NVIDIA GPU.
Open on GitHub → Explain another repo

← anil-matcha on gitmyhub — every repo by this author, as a profile.

Verify against the repo before relying on details.