Introduction
"When AI assistants move from the cloud to local, from a single model to a multi-model ecosystem, the true productivity revolution has only just begun."
This is Part 6 of the "Open Source Project of the Day" series. Today we explore AionUi (GitHub).
If you're looking for a free, open-source, cross-platform AI office assistant — one that can replace Claude Cowork but with even more features — AionUi is absolutely worth exploring. It supports multiple models, multiple sessions, WebUI remote access, file management, real-time preview, and AI image generation. It's a true "24/7 AI office partner."
Why this project?
- 🆓 Completely free and open source: Apache-2.0 license, no subscription fees
- 🌐 Cross-platform support: macOS, Windows, and Linux
- 🤖 Multi-model ecosystem: Supports Gemini, Claude, OpenAI, Qwen, Ollama, and more
- 🌐 WebUI remote access: 24/7 availability, supports mobile devices
- 📁 Smart file management: AI-driven file organization and processing
- 📄 Real-time preview: Instant preview for 9+ file formats
- 🎨 AI image generation: Integrated with multiple image generation models
What You'll Learn
- AionUi's core features and architectural design
- How to configure and use a multi-model AI assistant
- Setting up and using WebUI remote access
- Tips for file management and preview features
- AI image generation and editing capabilities
- Comparative analysis with Claude Cowork
- How to get started with AionUi quickly
Prerequisites
- Basic understanding of AI assistants
- Familiarity with desktop applications
- Understanding of API Key configuration (optional)
- Basic knowledge of file management
Project Background
Project Introduction
AionUi is a free, local, open-source 24/7 AI office assistant that supports multiple AI agents including Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, and Auggie. It provides a complete GUI interface and WebUI remote access, making the AI assistant a true "24-hour office partner."
Core problems the project solves:
- Claude Cowork only supports macOS and is locked to the Claude model
- Lack of cross-platform AI office assistant solutions
- Need for a unified interface supporting multiple models
- Need for local deployment to protect data privacy
- Need for remote access capability to enable 24/7 availability
Target user groups:
- Developers needing a cross-platform AI assistant
- Users looking to replace Claude Cowork
- AI users needing multi-model support
- Users with data privacy requirements
- Users who need remote access to AI tools
Author/Team Introduction
Team: iOfficeAI
- Background: Open-source team focused on AI office automation
- Philosophy: Build a free, open-source, cross-platform AI office assistant
- Project positioning: Open-source alternative to Claude Cowork
Project creation date: 2024 (based on GitHub activity, an actively maintained project)
Project Stats
- ⭐ GitHub Stars: 11.7k+ (rapidly and continuously growing)
- 🍴 Forks: 884+
- 📦 Version: v1.7.8 (latest version, released January 28, 2026)
- 📄 License: Apache-2.0 (fully open source, free to use)
- 🌐 Website: www.aionui.com
- 📚 Documentation: Includes complete usage guides and configuration docs
- 💬 Community: Active GitHub Issues and Discussions
- 👥 Contributors: 31 contributors, active community participation
Project development history:
- 2024: Project created, started building core features
- Mid-2024: Added multi-model support and WebUI features
- Late 2024: Refined file management and preview features
- 2025: Added AI image generation and editing features
- 2026: Continuous optimization, growing community activity
Main Features
Core Purpose
AionUi's core purpose is to provide a free, open-source, cross-platform AI office assistant, with main features including:
- Unified multi-model management: Supports Gemini, Claude, OpenAI, Qwen, Ollama, and more — unified conversation history management
- Cross-platform use: Full support for macOS, Windows, and Linux, with WebUI remote access
- Smart file management: AI-driven file organization, batch processing, Excel analysis, and more
- Real-time preview: Instant preview and editing for 9+ formats including PDF, Word, Excel, PPT, code, and Markdown
- AI image generation: Integrated with multiple image generation models including Gemini 2.5 Flash, Nano, and Banana
Quick Start
Installation
AionUi supports multiple installation methods:
# Method 1: Download from GitHub Releases (recommended)
# Visit https://github.com/iOfficeAI/AionUi/releases
# Download the installer for your platform
# macOS
# Download the .dmg file and install
# Windows
# Download the .exe file and install
# Linux
# Download the .AppImage or .deb/.rpm file and install
# Method 2: Build from source
git clone https://github.com/iOfficeAI/AionUi.git
cd AionUi
npm install
npm run build
Basic Configuration
1. First launch configuration
When starting AionUi for the first time, you need to configure an AI service:
# Supports Google account login (auto-links to Gemini CLI)
# Or configure using an API Key
# Gemini configuration example
provider: "gemini"
apiKey: "YOUR_GEMINI_API_KEY"
# Claude configuration example
provider: "anthropic"
apiKey: "YOUR_CLAUDE_API_KEY"
# OpenAI configuration example
provider: "openai"
apiKey: "YOUR_OPENAI_API_KEY"
# Local model configuration (Ollama)
provider: "custom"
apiUrl: "http://localhost:11434/v1"
2. Simplest usage example
# 1. Launch the AionUi application
# 2. Select an AI model (e.g., Gemini CLI)
# 3. Start chatting
User: "Help me organize the files on my desktop"
AI: "I'll help you organize your files. Please tell me your preferences..."
User: "Generate code for a React component"
AI: "I'll generate a React component for you..."
Core Features
- Multi-session management: Supports concurrent sessions, each with independent context, stored locally in SQLite
- WebUI remote access: Browser-based access, supports mobile devices, 24/7 availability, supports SSH tunneling
- Multi-model ecosystem: Major models (Gemini, OpenAI, Claude, Qwen) + local models (Ollama, LM Studio) + custom platforms
- Smart file management: File tree browsing, drag-and-drop upload, AI-powered smart organization, batch processing
- Real-time preview and editing: 9+ formats supported, real-time file change tracking, supports Markdown/code/HTML editing
- AI image generation: Integrated with multiple image generation models, supports image recognition and editing
- Interface customization: CSS-based interface customization, freely adjust colors, styles, and layout
- Data security: Fully local storage, data never goes to the cloud, user retains complete control
Project Advantages
| Comparison | AionUi | Claude Cowork | Other AI Assistants |
|---|---|---|---|
| Platform support | macOS / Windows / Linux | macOS Only | Partially cross-platform |
| Model support | Multi-model (Gemini, Claude, OpenAI, etc.) | Claude Only | Single model |
| Access method | GUI + WebUI remote access | GUI Only | Primarily GUI |
| Cost | Completely free and open source | $100/month subscription | Partially free |
| Data privacy | Fully local storage | Cloud storage | Hybrid storage |
| File management | AI-powered management | Basic support | Limited support |
| Preview features | 9+ formats real-time preview | Limited preview | Basic preview |
| Image generation | Multi-model support | Not supported | Partial support |
| Customization | CSS interface customization | Limited customization | Limited customization |
Why choose AionUi?
Compared to Claude Cowork, AionUi provides cross-platform support, a multi-model ecosystem, WebUI remote access, and is completely free and open source — making it a more powerful alternative.
Detailed Project Analysis
Architecture Design
AionUi uses a modern desktop application architecture based on Electron + React + TypeScript, consisting of several core modules:
Core Architecture
AionUi Application
├── Main Process (Electron)
│ ├── Window management
│ ├── System integration
│ ├── File system access
│ └── Local database management
├── Renderer Process (React)
│ ├── UI component layer
│ ├── State management
│ ├── API call layer
│ └── File preview components
├── AI Service Layer
│ ├── Multi-model adapters
│ ├── API call wrappers
│ ├── Streaming response handling
│ └── Context management
└── Data Layer
├── SQLite database (conversation history)
├── Local file storage
└── Configuration management
Multi-Model Adapter Architecture
AionUi supports multiple AI models through a unified adapter pattern. Each model (Gemini, Claude, OpenAI, etc.) has its own adapter implementation, managed through a unified AIModelAdapter interface. This design makes adding new models simple, with independent implementations for each model, supporting dynamic loading and configuration.
WebUI Remote Access Architecture
AionUi's WebUI is implemented through a built-in Express server and WebSocket, providing static file serving, API routes (chat, sessions, model management), and real-time communication. Supports security mechanisms including Token authentication, HTTPS, IP whitelisting, and SSH tunneling.
File Management System
AionUi's file management system supports file tree browsing, drag-and-drop upload, and AI-driven smart organization. The FileTreeManager scans the file system, uses AI to analyze files and execute organization operations. Supports recursive processing of files and folders.
Preview Panel System
AionUi's preview panel implements a plugin-based architecture through the PreviewComponent interface, supporting 9+ formats including PDF, Word, Excel, PPT, code, Markdown, HTML, images, and JSON/YAML/CSV. Each preview component implements canPreview, render, and update methods, with PreviewManager handling management and routing. Supports real-time file monitoring and editing.
Multi-Session Management System
AionUi's ConversationManager manages multiple concurrent sessions, where each session contains independent conversation history, model configuration, and metadata. Supports automatic title generation, intelligent context truncation (based on token limits), SQLite local persistence, and fast switching.
AI Image Generation System
AionUi implements a plugin-based image generation architecture through the ImageGenerator interface, supporting models including Gemini 2.5 Flash Image Preview, Nano, and Banana. The ImageGenerationManager uniformly manages all generators, supporting single and batch generation.
Local Data Security
AionUi uses a SQLite database to store all data (conversation history, file metadata, settings), with data fully stored in the user's data directory and never uploaded to any server. Supports optional data encryption and data export functionality, giving users complete control over their data.
Project Resources
Official Resources
- 🌟 GitHub: https://github.com/iOfficeAI/AionUi
- 📚 Documentation: GitHub Wiki
- 💬 Community: GitHub Discussions
- 🐛 Issue Tracker: GitHub Issues
- 📦 Latest version: v1.7.8 (released January 28, 2026)
Who Should Use This
AionUi is especially suitable for: Developers needing a cross-platform AI assistant, multi-model users, users with data privacy requirements, users who need office automation features, and those looking for a free open-source alternative to Claude Cowork.
Not suitable for: Users who only need simple chat, macOS users who don't need cross-platform support, users who don't need multi-model support.
Welcome to visit my personal homepage for more useful knowledge and interesting products

Top comments (0)