Spotify for AI — Switch between 7 AI models instantly
Your conversations, your rules. All data stays local, encrypted, and under your control.
Quick Start • Why Aether • Features • Architecture • Contributing
Like Spotify lets you switch between songs, Aether lets you switch between AI brains.
Pick the best AI for each task:
- 📝 Claude for writing and analysis
- 💻 GPT-4 for code generation
- 📄 Gemini for long documents (1M context)
- 🇨🇳 GLM/MiniMax for Chinese content
Your AI learns from every interaction:
- Automatic improvement: AI tracks performance and suggests optimizations
- User feedback: Rate responses (1-5 stars), AI adapts to your preferences
- Smart statistics: See success rates, response times, and satisfaction trends
- Learning reports: Comprehensive analytics on AI performance
Say what you need, AI creates it:
- "I need a tool to batch rename files" → Tool created instantly
- 7-language intent detection: Works in Chinese and English
- Auto-registration: New tools immediately available
- Safe execution: Generated tools are validated before use
Complex tasks, automated:
- 4 built-in templates: Code deployment, data processing, batch operations, notifications
- Cron scheduling: Run workflows on a schedule
- Smart control: Conditions, loops, parallel execution, auto-retry
- AI-powered: AI can trigger workflows intelligently
Production-grade infrastructure:
- API Gateway: JWT auth + rate limiting + unified REST API
- Real-time collaboration: Multi-user WebSocket sessions (like Google Docs)
- Plugin ecosystem: Load community extensions with security validation
- 40+ new APIs: Complete programmatic control
- 🏠 Local-first — Your conversations stay on your machine, encrypted, never sent to our servers
- 🔌 No vendor lock-in — Switch between 7 AI providers with one click, compare responses side-by-side
- 🛠️ Batteries included — 105+ tools ready to use: Gmail, GitHub, Google Workspace, Office automation
✅ Developers who need AI for code + docs + communication
✅ Power users tired of ChatGPT rate limits
✅ Teams who want full control over their AI stack
✅ Privacy-conscious users who don't trust cloud storage
❌ Casual users who only use ChatGPT web version occasionally
|
Ask the same question to Claude, GPT-4, and Gemini. Pick the best answer. Example: "Explain quantum computing"
|
Use the right AI for each job. Code review → GPT-4 (best at code) |
|
All conversations stored locally, encrypted with AES-256-GCM.
|
105+ built-in tools:
|
- Download Aether for your OS (Windows, macOS, Linux)
- Launch the app
- Add your API key (Claude/OpenAI/Gemini)
- Start chatting 🎉
Requires: Node.js 20+
git clone https://github.com/AnxForever/Aether.git
cd Aether
npm install
npm run build
npm start📚 See Quick Start Guide for detailed setup.
| Provider | Models | Special Features |
|---|---|---|
| Claude | Opus, Sonnet, Haiku | 200K context, vision |
| OpenAI | GPT-4, GPT-4 Turbo | Function calling, DALL·E |
| Gemini | Pro, Flash | 1M context, multimodal |
| MiniMax | abab6.5 | Chinese-optimized |
| Moonshot | v1 | 128K context |
| GLM | GLM-4 | Chinese NLP |
| DeepSeek | V2 | Cost-effective |
| 📧 Gmail 13 tools: Read, send, search, labels |
📊 Google Sheets 5 tools: Read, write, formulas |
| 💻 GitHub 6 tools: Repos, issues, PRs |
📄 Office 8 tools: PDF, Excel, PowerPoint |
| 📅 Calendar 5 tools: Events, scheduling |
📝 Google Docs 5 tools: Create, edit, format |
- AES-256-GCM encryption for sensitive data
- Scrypt key derivation (no hardcoded keys)
- Local storage — conversations never leave your machine
- Zero telemetry — we don't track you
- Input validation with Zod schemas
- OpenTelemetry distributed tracing
- Sentry error tracking & performance monitoring
- Real-time metrics dashboard
- Slack integration for team collaboration
- Dual Modes: Chat (creative) vs Coding (precise)
- Hot Reload: Auto-reload plugins without restart
- i18n: English + 简体中文
- Image Processing: HEIC, JPEG, PNG, WebP conversion
- Scheduled Tasks: Cron-based automation
- Self-Learning: Improve from user feedback
┌─────────────────────────────────────────┐
│ Electron Desktop App │
├─────────────────────────────────────────┤
│ Chat Mode │ Coding Mode │ Skills │
├─────────────────────────────────────────┤
│ Core Orchestration Engine │
│ Pipeline → AI Provider → Tool Execute │
├─────────────────────────────────────────┤
│ Claude │ GPT-4 │ Gemini │ ... │
├─────────────────────────────────────────┤
│ Storage │ Plugins │ Monitoring │ i18n │
└─────────────────────────────────────────┘
Key Components:
- Orchestrator: Manages conversation cycles
- Pipeline: Context → AI → Tools → Response
- Connectors: 7 AI providers with streaming
- Skills: 105+ tools across 8 categories
- Storage: SQLite with encryption
- Quick Start Guide — Get started in 5 minutes
- Architecture Overview — System design deep dive
- API Reference — Complete API documentation
We welcome contributions! See CONTRIBUTING.md for guidelines.
Ways to contribute:
- 🐛 Report bugs
- ✨ Request features
- 🔧 Submit PRs
- 📖 Improve documentation
- 19,302 lines of TypeScript
- 7 AI providers integrated
- 105+ built-in tools
- 100% type safety (strict mode)
This project is licensed under the MIT License - see the LICENSE file for details.
Made with ❤️ by AnxForever
Star this repo if Aether helps you! ⭐




