VGO CODE is a professional desktop agent workspace designed for advanced conversational AI workflows, built on Electron with a React + Vite rendering layer.
VGO CODE delivers a polished desktop application experience for managing multiple AI sessions, switching between model engines, and integrating remote AI backends. Its architecture is optimized for desktop deployment and includes a local API template service for rapid extension.
- Fully packaged Electron desktop application
- Multi-session chat management
- Multiple engine selection with VGO Remote support
- VGO AI account binding and model selection
- Local API service template in
server/ - Production-ready desktop installer distribution
Download the latest installer from:
For registration, deployment, or product updates, visit:
E:\VGO-CODE
├─ build/ # Electron build resources
├─ docs/ # Project documentation
├─ electron/ # Electron main process and preload logic
│ ├─ main.js
│ ├─ preload.js
│ └─ core/
├─ src/ # Frontend source code
├─ dist-web/ # Built frontend assets for Electron rendering
├─ server/ # Local API and integration service templates
├─ scripts/ # Utility and release scripts
├─ vendor/ # Embedded third-party libraries
├─ package.json # Project metadata and scripts
└─ README.md # Project overview and instructions
The Electron main process loads the single React/Vite renderer bundle at dist-web/index.html. The old static UI fallback has been removed to keep one production UI path.
Start the application in development mode:
npm startBuild the frontend assets:
npm run build:webCreate directory-based Electron package:
npm run packGenerate the installer package:
npm run distStart the API server independently from the repository root:
cd E:\VGO-CODE\server
npm startCommon endpoints:
GET /healthGET /modelsPOST /auth/registerPOST /auth/loginPOST /chat
docs/ARCHITECTURE.mddocs/VGO-CORE-ROADMAP.md
For guided setup, installer downloads, and release notes, visit: