Transform your command-line experience into a modern, efficient AI Chat interface.
English | 简体中文 | 日本語 |Official Site | Twitter
Seamlessly integrate multiple terminal AI agents - Gemini CLI, Claude Code, Qwen Code and more
Intelligent image generation, editing, and recognition powered by Gemini 2.5 Flash Image Preview - the most advanced image model, with support for other leading AI image models
💡 Need help setting up free image generation? Follow the tutorial to configure image generation models - Complete step-by-step guide for OpenRouter setup and free Gemini 2.5 Flash Image Preview configuration.
Batch renaming, auto organization, smart classification, file merging
AI helps you create, organize, analyze, and beautify Excel files
Multiple conversations, no task confusion, independent memory, double efficiency
This is just the tip of AionUi's capabilities! 🚀
Want to explore more features? Keep reading to discover what else AionUi can help you with:
- 🎯 Write code, create documents, analyze data
- 🗂️ Learn new things, answer questions, translate text
- ⚡ And many more daily work and learning scenarios
While the official Gemini CLI is powerful, its command-line interface has limitations for daily use. AionUi provides a GUI alternative that addresses these key pain points:
- Using the
@command to select files is cumbersome- Conversations are lost when closing the CLI window
- Command-line interface lacks natural chat interactions
- Single conversation mode limits parallel workflows
- Restricted to Gemini models only, unable to use other excellent large language models
AionUi provides a modern interface for users who need better workflow efficiency, while breaking the single-model limitation, allowing you to choose the most suitable AI model for different task requirements.
- Multi-Conversation - Open multiple chats simultaneously, no interference
- Permanent Storage - All conversations saved locally, never lost
- Modern Interface - Chat interface like WeChat, simple to use
- Multi-Model Support - Not just Gemini, can use other AI models too
- File Tree Browsing - Browse files like folders, click to use
- File Upload - Drag and drop files, AI helps you process
- Code Comparison - File before/after comparison, clear at a glance
- Smart Organization - AI helps organize folders, automatic classification
- Excel Processing - AI helps create and modify Excel files
- Function Calling - Complete Gemini API, more powerful features
- Code Rendering - Code blocks display more beautifully, format clearer
- Tool Scheduling - Automatically select most suitable tools, no manual selection needed
- Multi-Terminal Agent Support - Seamlessly integrate with various terminal AI agents (Gemini CLI, Claude Code, Qwen Code, etc.)
- Dynamic CLI Detection - Automatic backend discovery and CLI path detection
- Secure Authentication - OAuth support and secure authentication flows
- Real-time Monitoring - Live connection status and backend health monitoring
- Unified Interface - All terminal AI agents accessible through the same chat interface
- Multi-Key Rotation - Automatic rotation of multiple API keys for enhanced reliability
- Intelligent Error Recovery - Time-based blacklisting (90s) with automatic retry mechanisms
- High Availability - Seamless failover between API keys to prevent service interruptions
- Rate Limit Handling - Smart retry logic that respects API rate limits and quotas
- Performance Optimization - Reduced downtime and improved API call success rates
- Smart Image Generation - Powered by Gemini 2.5 Flash Image Preview, the most advanced image model
- Multi-Model Support - Also supports other leading AI image models for diverse creative needs
- Intelligent Editing - AI-powered image editing and enhancement capabilities
- Image Recognition - Advanced image analysis and understanding
- High-Quality Output - Professional-grade image generation with detailed control
- Multi-Platform Support - Support Gemini, OpenAI, ModelScope, OpenRouter, etc.
- Flexible Configuration - Each platform can configure multiple models, support custom addresses
- Easy Login - Support Google account login, no need to remember API keys
- Auto Fix - Automatically detect and fix configuration issues, no manual debugging needed
Ready to try AionUi? Download the latest version for your platform from our releases page:
- Node.js >= 16.0.0
- npm >= 8.0.0
- At least one AI service API key or authentication configuration:
- Gemini: Get Gemini API key
- OpenAI: Get OpenAI API key
- ModelScope: Get ModelScope API key
- OpenRouter: Get OpenRouter API key
- Terminal AI Agents: Gemini CLI, Claude Code, Qwen Code, etc.
-
Clone the repository
git clone https://github.com/iOfficeAI/AionUi.git cd AionUi -
Install dependencies
npm install
-
Configure Authentication
- Open the application and go to Settings
- Configure at least one AI service authentication:
- Gemini: API Key, Vertex AI, or OAuth Personal Authentication
- OpenAI: API Key or custom endpoints
- ModelScope: API Key
- OpenRouter: API Key
- Terminal AI Agents: Install and configure Claude Code, Qwen Code, etc.
-
Start the application
npm start
# Build macOS version
npm run build-mac --arch=arm64 # Apple Silicon
npm run build-mac --arch=x64 # Intel
# Build Windows version
npm run build-win
# Build all platforms
npm run build- Desktop App: Electron 37.2.0
- Frontend Framework: React 19.1.0
- UI Component Library: Arco Design Web React
- AI Engine: Google Gemini CLI Core
- Styling Framework: UnoCSS
- Build Tools: Webpack + TypeScript
- Icon Library: IconPark React
AionUI/
├── src/
│ ├── adapter/ # Adapter layer
│ ├── agent/ # AI agents
│ │ └── gemini/ # Gemini AI integration
│ ├── common/ # Common modules
│ ├── process/ # Main process
│ ├── renderer/ # Renderer process
│ │ ├── components/ # UI components
│ │ ├── conversation/# Conversation related
│ │ └── messages/ # Message handling
│ └── worker/ # Worker process
├── config/ # Configuration files
├── public/ # Static resources
└── package.json
- Code Development: Code review, refactoring suggestions, bug fixes
- Document Writing: Automatic document generation, report summaries
- Data Analysis: Data visualization, analysis reports
- Project Management: Task planning, progress tracking
- Learning Assistant: Knowledge Q&A, concept explanation
- Daily Office Work: Email writing, meeting notes, work summaries
- Learning & Growth: Language learning, skill training, knowledge organization
- Creative Work: Copywriting, brainstorming, inspiration collection
- AI Image Generation: Create, edit, and enhance images with AI-powered tools
- Multi-Model Collaboration: Choose the most suitable AI model based on task characteristics
- Gemini: Code generation, technical documentation, image generation & editing
- OpenAI: Creative writing, content creation
- ModelScope: Chinese understanding, localization tasks
- OpenRouter: Cost optimization, model comparison
- Multi-Terminal Agent Workflows: Leverage different terminal AI agents for specialized tasks
- Unified Integration: Seamless integration with Gemini CLI, Claude Code, Qwen Code, and other terminal AI agents
- High Availability: Automatic failover between multiple API keys for uninterrupted service
Supports multiple authentication methods and platforms:
-
Gemini Platform:
- Gemini API Key: Direct use of Gemini API
- Vertex AI: Use Google Cloud Vertex AI
- Personal Authentication: OAuth personal authentication
-
Other Platforms:
- OpenAI Compatible: Support any service compatible with OpenAI API
- ModelScope: Support Alibaba Cloud ModelScope platform
- OpenRouter: Support OpenRouter aggregation platform
- Custom Platform: Support custom API endpoints and models
-
Multi-API Key Configuration:
- Key Rotation: Configure multiple API keys for automatic rotation
- Error Recovery: 90-second blacklisting with automatic retry
- High Availability: Seamless failover between available keys
- Rate Limit Handling: Smart retry logic for optimal performance
Integration with external terminal AI agents:
- Gemini CLI: Built-in support for Gemini CLI (AionUi's core)
- Claude Code: Support for Claude Code terminal agent
- Qwen Code: Support for Qwen Code terminal agent
- Dynamic Detection: Automatic discovery of available terminal AI agents
Supports HTTP proxy configuration for network-restricted environments.
We envision AionUi evolving into a Universal AI Agent Platform that democratizes powerful AI agents for everyday users:
- Terminal Assistants: Starting with Gemini CLI, will support more terminal tools in the future
- Browser Assistants: Integrate open-source web automation tools to help with web tasks
- Unified Interface: All AI assistants use the same simple chat interface
- Easy Discovery: New AI assistants can be easily found and used
- Multi-Model Support: Can use various AI models like Gemini, Claude, GPT, etc.
- Switch Anytime: Use whichever model you want without changing your workflow
- Independent Configuration: Each model has independent settings, no interference
- Smart Recommendations: Automatically recommend the most suitable AI model based on task
- Cost Comparison: Help you choose the most cost-effective model
Our goal is to make AI agents simple and easy to use, so ordinary users can get started easily. We believe that:
- Simplicity is beauty: Complex AI features should be simple to use
- AI should understand users: Users shouldn't need to adapt to AI, but AI should adapt to users
- Open source is more transparent: We prioritize using open-source agents so everyone can see the code
- Chat is most natural: Complex work can be done through simple chat
AionUi aims to bridge the gap between powerful AI capabilities and everyday usability, making sophisticated AI agents as easy to use as chatting with a friend.
This project is licensed under the Apache-2.0 License.
Issues and Pull Requests are welcome!
- Fork this project
- Create a feature branch (
git checkout -b feature/AmazingFeature) - Commit your changes (
git commit -m 'Add some AmazingFeature') - Push to the branch (
git push origin feature/AmazingFeature) - Open a Pull Request
⭐ Star the repo if you like it