Documentation Index
Fetch the complete documentation index at: https://mintlify.com/anil-matcha/open-generative-ai/llms.txt
Use this file to discover all available pages before exploring further.
Open Generative AI offers three deployment paths to match different use cases. Download the desktop app for a no-setup, one-click install on macOS, Windows, or Linux. Use Docker to run a containerized production build on any server or home lab. Or build from source to produce custom production binaries or run the Next.js web server directly — useful for contributors, CI pipelines, and environments where you want full control over the build.
Desktop App
Docker
From Source
The desktop app is the recommended option for individual users. It requires no Node.js, no terminal, and no configuration — just download, install, and generate.Downloads
| Platform | Download |
|---|
| macOS Apple Silicon (M1/M2/M3/M4) | Open Generative AI-1.0.9-arm64.dmg |
| macOS Intel (x64) | Open Generative AI-1.0.9.dmg |
| Windows (x64) | Open Generative AI Setup 1.0.9.exe |
| Linux (Ubuntu x64) | .AppImage / .deb from the v1.0.9 release |
macOS Installation
Because the app is not notarized by Apple, macOS Gatekeeper will block it on first launch. Follow these steps to open it:Step 1 — Mount the .dmg and drag Open Generative AI to /Applications.Step 2 — Open Terminal and run:xattr -cr "/Applications/Open Generative AI.app"
/Applications → click Open → click Open again on the dialog.You only need to do this once. After the first launch the app opens normally like any other application.
- Try to open the app — macOS will show a blocked dialog.
- Go to System Settings → Privacy & Security.
- Scroll down to find “Open Generative AI was blocked”.
- Click Open Anyway → Open.
Windows Installation
Windows SmartScreen may display a warning because the installer is not code-signed.
- Click More info on the SmartScreen dialog.
- Click Run anyway.
The app installs silently to %LocalAppData% and adds a Start Menu shortcut.
Linux Installation
AppImage
Make the AppImage executable and run it directly:chmod +x "Open Generative AI-1.0.9.AppImage"
./"Open Generative AI-1.0.9.AppImage"
libfuse2:sudo apt install libfuse2
.deb Package
Install the .deb on Debian or Ubuntu:sudo apt install ./open-generative-ai_1.0.9_amd64.deb
Ubuntu 24.04+ — AppArmor sandbox restriction
Ubuntu 24.04 and later enforce a kernel security policy (apparmor_restrict_unprivileged_userns) that blocks Chromium’s user-namespace sandbox. If the app fails to start silently or crashes immediately, use one of the following fixes:Option B makes a system-wide change. Prefer Option A (the .deb package) whenever possible — it ships an AppArmor profile that grants the required permission automatically without any system-wide modification.
.deb package instead of the AppImage. The .deb ships with an AppArmor profile that grants the required permission on install automatically.Option B — Temporary fix (AppImage users):sudo sysctl -w kernel.apparmor_restrict_unprivileged_userns=0
echo 'kernel.apparmor_restrict_unprivileged_userns=0' | sudo tee /etc/sysctl.d/99-userns.conf
Docker is the fastest path to a production-ready web deployment. The included docker-compose.yml builds the app from source and serves it on port 3001.docker-compose.yml
services:
open-generative-ai:
build: .
container_name: open-generative-ai
ports:
- "3001:3000"
environment:
- NODE_ENV=production
restart: unless-stopped
3000 internally and maps it to 3001 on the host. The restart: unless-stopped policy keeps the service running after system reboots.Start the container
From the repo root (the directory containing docker-compose.yml):The first build compiles all workspace packages and runs next build, so it will take a few minutes. Subsequent starts are fast because Docker caches the build layers.Once running, open your browser at http://localhost:3001.Environment variables
The Muapi API key is entered through the app’s UI on first use and stored in the browser’s localStorage — no server-side environment variable is required for normal operation.If you want to pre-configure the key for a shared or automated deployment, pass it as an environment variable in docker-compose.yml:environment:
- NODE_ENV=production
- MUAPI_API_KEY=your_key_here
Stop the container
Building from source gives you full control over the output and is required if you want to customize the codebase before deploying. Make sure you have completed the Quickstart setup steps (cloned with submodules and run npm run setup) before proceeding.Production Web Build
Build the optimized Next.js bundle and start the production server:npm run build
npm run start
3000 by default.
Desktop App Builds
Use Electron Builder to produce native installers. All outputs are written to the release/ directory.macOS (DMG — Intel + Apple Silicon)
Produces two .dmg files in release/: one for arm64 (Apple Silicon) and one for x64 (Intel).Windows (NSIS installer — x64)
npm run electron:build:win
.exe installer in release/. See the Desktop App tab for the SmartScreen bypass instructions end users will need.Linux (AppImage + .deb — x64)
npm run electron:build:linux
.AppImage and a .deb package in release/. See the Desktop App tab for installation and AppArmor notes.npm run electron:build:all
Cross-platform builds (e.g., building a Windows installer from macOS) require the target platform’s toolchain to be available. For reliable cross-platform builds, use a CI runner that matches the target OS.
