Socials:
The Open Source Alternative to Cluely - A lightning-fast, privacy-first AI assistant that works seamlessly during meetings, interviews, and conversations without anyone knowing.
This is the open source version of the $15M company Cluely 🎯. Experience the same powerful real-time AI assistance, but with complete transparency, privacy, and customization control.
Available formats: .dmg (macOS) • .msi (Windows) • .exe (Windows) • .deb (Linux) • .rpm (Linux) • .AppImage (Linux)
The world's most efficient AI assistant that lives on your desktop
| 🪶 Ultra Lightweight | 📺 Always Visible | ⚡ Instant Access |
|---|---|---|
| Only ~10MB total app size | Translucent overlay on any window | One click to activate AI assistance |
| 27x smaller than Cluely (~270MB) | Always on top, never intrusive | Overlaps seamlessly with your workflow |
| 50% less compute power usage | Perfect transparency level | Ready when you need it most |
| Feature | 🟢 Pluely (Open Source) | 🔴 Original Cluely |
|---|---|---|
| App Size | ~10MB ⚡ | ~270MB 🐌 |
| Size Difference | 27x Smaller 🪶 | Bloated with unnecessary overhead |
| Compute Usage | 50% Less CPU/RAM 💚 | Heavy resource consumption |
| Startup Time | <100ms ⚡ | Several seconds |
| Privacy | 100% Local with your LLM 🔒 | Data sent to servers |
| Cost | Free & Open Source 💝 | $15M company pricing 💸 |
Pluely comes with powerful global keyboard shortcuts that work from anywhere on your system:
| Shortcut | macOS | Windows/Linux | Function |
|---|---|---|---|
| Toggle Window | Cmd + \ |
Ctrl + \ |
Show/Hide the main window + app icon (based on settings) |
| Voice Input | Cmd + Shift + A |
Ctrl + Shift + A |
Start voice recording |
| Screenshot | Cmd + Shift + S |
Ctrl + Shift + S |
Capture screenshot |
| System Audio | Cmd + Shift + M |
Ctrl + Shift + M |
Toggle system audio capture |
Control whether the Pluely window stays above all other applications for instant access:
- Enabled: : Window always appears on top of other applications
- Disabled: Window behaves like normal applications
- Settings Control: Toggle in Settings → Always On Top Mode
- Independent Control:: Works separately from stealth features
- Perfect for: Quick access during meetings, presentations, or when you need instant AI assistance
Perfect for instant access during sensitive meetings and conversations - window stays above all applications for seamless AI assistance!
Complete stealth mode control integrated with main toggle:
- Show Mode (Default): App icon remains visible in dock/taskbar when window is hidden
- Hide Mode: App icon completely disappears from dock/taskbar when window is hidden (app keeps running in background)
- Settings Control: Configure dock/taskbar icon visibility in Settings → App Icon Visibility
- Auto-Integration: Works automatically with main toggle shortcut (
Cmd+\/Ctrl+\) based on your settings - Cross-Platform: Works seamlessly on macOS (ActivationPolicy::Accessory), Windows (skip_taskbar), and Linux (skip_taskbar)
Perfect for maximum discretion during sensitive meetings and conversations - the app icon disappears from dock/taskbar but app remains fully functional!
Transform any audio on your system into an intelligent AI assistant with real-time transcription and contextual help:
- Real-Time AI Assistant: 8 specialized modes for meetings, interviews, presentations, learning, and more
- Background Processing: Captures system audio continuously with smart context detection
- Multi-Scenario Support: Meeting insights, interview help, translation, presentation coaching, and learning assistance
- Cross-Platform: Works on macOS, Windows, and Linux with platform-specific audio routing
- Settings Control: Toggle system audio capture in Settings → Audio section
- AI Integration: Seamlessly connects transcription to AI models for instant responses
Turn any audio into your personal AI assistant - from Zoom calls to podcasts, get intelligent help in real-time!
🔒 Security by Design: Operating systems block direct system audio access for privacy protection - this is universal across all platforms and applications.
⚙️ Quick Setup Process:
- macOS: Install BlackHole (free) + Audio MIDI Setup (5 min)
- Windows: Enable built-in Stereo Mix OR install VB-Cable (free) (3 min)
- Linux: Use built-in PulseAudio monitors (usually works out-of-box)
📖 Complete Audio Setup Guide • Step-by-step for all platforms
Experience multiple input methods including text entry, and image pasting:
| Input Example 1 | Input Example 2 |
|---|---|
 |
 |
| Clean, intuitive text interface with markdown support for seamless conversation | Image pasting functionality that automatically adds files to attachments for AI analysis |
Capture and analyze screenshots with automatic or manual modes for different use cases:
| Manual Mode Screenshot |
|---|
 |
| Screenshots are captured and automatically added to your attached files for manual submission with your own prompt |
| Auto Mode Screenshot |
|---|
 |
| Screenshots are automatically submitted to AI using your custom prompt for instant analysis |
Speak naturally and get instant AI responses with advanced voice activity detection:
Real-time speech-to-text using OpenAI Whisper with automatic voice detection - just speak and get AI responses
Keep track of all your AI interactions with persistent conversation history:
Browse through previous conversations, search chat history, and continue where you left off
Get started with your preferred AI provider in seconds. Pluely supports all major LLM providers with dynamic model fetching, custom providers, and seamless integration:
| Custom Provider Setup | OpenAI Setup |
|---|---|
 |
 |
| Add your custom provider and select the provider from AI providers in the AI Providers section, then enter your API key and AI model | Add your OpenAI API key and select from dynamically fetched models including GPT-4, GPT-4 Turbo, and more |
| Google Gemini Setup | xAI Grok Setup |
|---|---|
 |
 |
| Connect to Google Gemini with your API key and choose from available models like Gemini Pro | Integrate with xAI Grok - dynamically fetches available models for seamless AI assistance |
| Anthropic Claude Setup | Additional AI Providers |
|---|---|
 |
🔸 Mistral AI 🔸 Groq 🔸 Cohere 🔸 Perplexity AI |
| Enter your Claude API key and model name manually for powerful AI conversations | These providers are also available in the AI Providers settings for more options |
Each provider comes pre-configured with optimal settings:
- 🔄 Dynamic Model Fetching: Automatically fetch latest models from OpenAI, xAI, Gemini, Mistral, and Groq
- 🎯 Manual Model Selection: For Claude and Cohere, enter your preferred model name
- 🔐 Secure Authentication: API keys stored locally in browser storage
- 📡 Streaming Support: Real-time responses for all providers
- 🖼️ Multi-Modal: Image analysis support where available
- ⚡ Fast Integration: One-click setup for instant AI assistance
- Choose Your Provider: Select from the comprehensive list in Settings
- Enter API Key: Add your provider's API key securely
- Select Model: Choose from dynamically fetched models or enter manually
- Start Chatting: Begin your AI-powered conversations immediately
For providers not in our list, use the custom provider option:
- Add any OpenAI-compatible API endpoint
- Configure custom authentication methods
- Set up your own response parsing
- Full flexibility for any LLM service
Pluely supports advanced voice input with multiple speech-to-text providers for accurate, real-time transcription. Choose from industry-leading STT services or add your own custom provider:
| Provider | Auth Type | Features | Model |
|---|---|---|---|
| 🎤 OpenAI Whisper | Bearer Token | High accuracy, fast processing | whisper-1 |
| ⚡ Groq Whisper | Bearer Token | Ultra-fast transcription | whisper-large-v3 |
| 🔊 ElevenLabs STT | xi-api-key | Natural language processing | scribe_v1 |
| 🔍 Google Speech-to-Text | Bearer Token | Advanced language detection | N/A |
| 🎯 Deepgram STT | Bearer Token | Real-time processing | nova-2 |
| 🧠 IBM Watson STT | Basic API Key | Enterprise-grade accuracy | en-US_BroadbandModel |
| ☁️ Azure Speech-to-Text | Subscription Key | Cloud-native processing | N/A |
| 🎵 Speechmatics | Bearer Token | High accuracy for various languages | N/A |
| 🤖 Zhipu GLM-ASR | Bearer Token | Chinese language specialist | glm-asr |
| 🚀 Doubao STT | Bearer Token | Fast Chinese transcription | whisper-1 |
- 🎯 Real-time Processing: Instant speech recognition with voice activity detection
- 🌍 Multi-Language Support: Choose providers optimized for specific languages
- ⚡ Fast & Accurate: Industry-leading transcription accuracy and speed
- 🔒 Secure Authentication: API keys stored locally and securely
- 🎨 Seamless Integration: Works with Pluely's voice input features
Can't find your preferred speech-to-text provider? Add custom STT providers with full configuration control:
- 🔧 Flexible API Integration: Support for any REST API endpoint
- 🔐 Custom Authentication: Bearer tokens, API keys, or custom headers
- 📝 Request Customization: Configure audio formats, parameters, and headers
- 🎯 Response Parsing: Define custom response paths for transcription text
- ⚙️ Advanced Configuration: Set up audio formats, sample rates, and more
- Select Provider: Choose from the comprehensive STT provider list
- Configure API: Enter your provider's API credentials securely
- Test Connection: Verify your setup with a quick voice test
- Start Speaking: Use voice input seamlessly in your conversations
🎯 The Ultimate Stealth AI Assistant - Invisible to Everyone
Pluely is engineered to be completely invisible during your most sensitive moments:
- 🔍 Undetectable in video calls - Works seamlessly in Zoom, Google Meet, Microsoft Teams, Slack Huddles, and all other meeting platforms
- 📺 Invisible in screen shares - Your audience will never know you're using AI assistance
- 📸 Screenshot-proof design - Extremely difficult to capture in screenshots due to translucent overlay
- 🏢 Meeting room safe - Won't appear on projectors or shared screens
- 🎥 Recording invisible - Doesn't show up in meeting recordings or live streams
- ⚡ Instant hide/show - Overlaps with any window on your PC/Desktop for quick access
Perfect for confidential scenarios where discretion is absolutely critical.
Built with Tauri, Pluely delivers native desktop performance with minimal resource usage:
- Instant startup - launches in milliseconds
- 10x smaller than Electron apps (~10MB vs ~100MB)
- Native performance - no browser overhead
- Memory efficient - uses 50% less RAM than web-based alternatives
- Cross-platform - runs natively on macOS, Windows, and Linux
Unlike cloud-based solutions, Pluely keeps everything local:
- Your data never touches any servers - all processing happens locally and with your LLM provider
- API keys stored securely in your browser's localStorage
- No telemetry or tracking - your conversations stay private, not stored anywhere
- Offline-first design - works without internet (except for AI API calls)
Pluely sits quietly on your desktop, ready to assist instantly with zero setup time:
Completely undetectable in all these critical situations:
- Job interviews - Get real-time answers without anyone knowing you're using AI
- Sales calls - Access product information instantly while maintaining professionalism
- Technical meetings - Quick reference to documentation without breaking flow
- Educational presentations - Learning assistance that's invisible to your audience
- Client consultations - Professional knowledge at your fingertips, completely discrete
- Live coding sessions - Get syntax help, debug errors, explain algorithms without detection
- Design reviews - Analyze screenshots, get UI/UX suggestions invisibly
- Writing assistance - Grammar check, tone adjustment, content optimization in stealth mode
- Board meetings - Access information without anyone noticing
- Negotiations - Real-time strategy assistance that remains completely hidden
Connect to any AI provider with built-in support for:
- OpenAI - Dynamically fetches models directly from OpenAI, select your own
- Anthropic Claude - Enter your model name manually
- xAI Grok - Dynamically fetches models directly from xAI, select your own
- Google Gemini - Dynamically fetches models directly from Gemini, select your own
- Custom providers - Coming soon
Pluely offers two powerful audio input methods:
- Voice Activity Detection (VAD) using
@ricky0123/vad-react - Real-time voice commands - just speak and get AI responses
- Works immediately - no setup required on any platform
- Meeting transcription - capture Zoom, Teams, Google Meet audio
- Video analysis - transcribe YouTube, educational content, podcasts
- Real-time processing with automatic speech detection
- Cross-platform support - works on macOS, Windows, Linux
- Setup required - see System Audio Setup Guide
With any industry-leading STT providers: OpenAI Whisper, Groq, ElevenLabs, Google, Deepgram, and more!
- Text input with markdown support
- Image attachments - analyze images, screenshots
- Translucent window that overlays any application
- Always on top but never in the way
- Minimal design - focus on content, not interface
- Dark/light theme support
- App Icon Stealth Mode - Complete dock/taskbar invisibility for maximum discretion
- Intelligent Icon Management - Automatically hide/show based on window visibility settings
- Cross-Platform Stealth - Works seamlessly on macOS (Dock), Windows (Taskbar), and Linux (Panel)
- Global Keyboard Shortcuts - Control app from anywhere without detection
- Chunked streaming - see responses as they're generated
- Cancellable requests - stop generation anytime
- Error handling with retry mechanisms
Pluely works seamlessly across all major operating systems with comprehensive audio capture support.
Pluely offers multiple audio capture options to work with any setup:
- ✅ No setup required - works out of the box on all platforms
- ✅ Uses your microphone for voice input and commands
- ✅ Perfect for: Voice dictation, personal notes, voice commands
- ✅ Supported: macOS, Windows, Linux
- 🎯 Captures everything your computer plays - meetings, videos, music
- 🎯 Perfect for: Meeting transcription, video analysis, system audio monitoring
- ⚙️ Requires: Virtual audio device setup (platform-specific)
- ✅ Supported: macOS, Windows, Linux
| Platform | Default Input | System Audio | Virtual Audio Solutions |
|---|---|---|---|
| 🍎 macOS | ✅ Built-in | ✅ Supported | BlackHole (free), Loopback ($109), SoundFlower |
| 🪟 Windows | ✅ Built-in | ✅ Supported | Stereo Mix (built-in), VB-Audio Cable (free), VoiceMeeter (free) |
| 🐧 Linux | ✅ Built-in | ✅ Supported | PulseAudio Monitors (built-in), ALSA Loopback |
Once configured, system audio capture enables transcription of:
- 🎥 Meeting Audio: Zoom, Teams, Google Meet, Discord, Slack
- 🎬 Video Content: YouTube, Netflix, Twitch, educational videos
- 🎵 Music & Podcasts: Spotify, Apple Music, podcast apps
- 🎮 Game Audio: Game commentary, streaming content
- 📞 VoIP Calls: Skype, WhatsApp calls, phone calls
- 📱 Any System Audio: Everything your computer plays
- Default Input: ✅ Instant - no setup required
- System Audio: ⚙️ Easy - install BlackHole, configure Multi-Output
- Time: ~5 minutes for system audio setup
- Best Option: BlackHole (free) or Loopback (paid but easier)
- Default Input: ✅ Instant - no setup required
- System Audio: ⚙️ Easy - enable Stereo Mix or install VB-Cable
- Time: ~3 minutes for system audio setup
- Best Option: Stereo Mix (built-in) or VB-Audio Cable (free)
- Default Input: ✅ Instant - no setup required
- System Audio: ⚙️ Easy - PulseAudio monitors usually work out-of-box
- Time: ~2 minutes for system audio setup
- Best Option: PulseAudio Monitor devices (usually pre-configured)
Pluely includes comprehensive debugging tools for all platforms:
- 🔍 Device Debug: Shows all available audio devices with detailed analysis
- 🎵 Audio Level Testing: 3-second audio level test with real-time feedback
- 📊 Real-time Monitoring: Live RMS/Peak audio level display in console
- 🎯 Smart Device Detection: Automatically selects best available device
- 🔄 Auto-Recovery: Handles device conflicts and "already running" errors
- 📖 SYSTEM_AUDIO_SETUP.md: Complete setup guide for all platforms
- 🎯 Platform-specific instructions: Step-by-step guides with screenshots
- 🔧 Troubleshooting: Common issues and solutions for each OS
- 💡 Pro Tips: Best practices and optimization recommendations
- Download Pluely for your platform from releases
- Install & Launch the application
- Click the audio button (headphones icon) in the interface
- Choose your method:
- "🎤 Default Input" - works immediately, no setup
- "🔄 System Audio" - requires virtual audio setup (see guide)
- Start speaking or play audio - transcription begins automatically!
- 🌐 Universal Access: Use Pluely on any device, anywhere
- 🔄 Consistent Experience: Same features and UI across all platforms
- 🛠️ Flexible Setup: Multiple audio solutions to fit any environment
- 📈 Future-Proof: Built with Tauri for native performance on all platforms
- Node.js (v18 or higher)
- Rust (latest stable)
- npm or yarn
🎵 Audio Capabilities: The built app includes full cross-platform audio support with both microphone input and system audio capture. See SYSTEM_AUDIO_SETUP.md for platform-specific setup guides.
# Clone the repository
git clone https://github.com/iamsrikanthnani/pluely.git
cd pluely
# Install dependencies
npm install
# Start development server
npm run tauri dev# Build the application
npm run tauri buildThis creates platform-specific installers in src-tauri/target/release/bundle/:
- macOS:
.dmgfile - Windows:
.msiinstaller - Linux:
.debfile
- API keys are stored in browser localStorage (not sent to any server)
- No telemetry - your keys never leave your device
- Session-based - keys can be cleared anytime
- Direct HTTPS connections to AI providers
- No proxy servers - your requests go straight to the AI service
- Request signing handled locally
- TLS encryption for all API communications
- No conversation logging - messages aren't stored anywhere
- Temporary session data - cleared on app restart
- Local file processing - images processed in-browser
- No analytics - completely private usage
Pluely makes API calls directly from your frontend because:
- 🔒 Maximum Privacy: Your conversations never touch our servers
- 🚀 Better Performance: Direct connection = faster responses
- 📱 Always Local: Your data stays on your device, always
- 🔍 Transparent: You can inspect every network request in dev tools
- ⚡ No Bottlenecks: No server capacity limits or downtime
- Native binary - no JavaScript runtime overhead
- Rust backend - memory-safe, ultra-fast
- Small bundle size - ~10MB total app size
- Instant startup - launches in <100ms
- Low memory usage - typically <50MB RAM
- React 18 with concurrent features
- TypeScript for compile-time optimization
- Vite for lightning-fast development
- Tree shaking - only bundle used code
- Lazy loading - components load on demand
src/
├── components/ # Reusable UI components
│ ├── completion/ # AI completion interface
│ │ ├── Audio.tsx # Audio recording component
│ │ ├── AutoSpeechVad.tsx # Voice activity detection
│ │ ├── Files.tsx # File handling component
│ │ ├── index.tsx # Main completion component
│ │ └── Input.tsx # Text input component
│ ├── Header/ # Application header
│ ├── history/ # Chat history components
│ │ ├── ChatHistory.tsx # Chat history management
│ │ ├── index.tsx # History exports
│ │ └── MessageHistory.tsx # Message history display
│ ├── Markdown/ # Markdown rendering
│ ├── Selection/ # Text selection handling
│ ├── settings/ # Configuration components
│ │ ├── ai-configs/ # AI provider configurations
│ │ │ ├── CreateEditProvider.tsx # Custom AI provider setup
│ │ │ ├── CustomProvider.tsx # Custom provider component
│ │ │ ├── index.tsx # AI configs exports
│ │ │ └── Providers.tsx # AI providers list
│ │ ├── DeleteChats.tsx # Chat deletion functionality
│ │ ├── Disclaimer.tsx # Legal disclaimers
│ │ ├── index.tsx # Settings panel
│ │ ├── ScreenshotConfigs.tsx # Screenshot configuration
│ │ ├── stt-configs/ # Speech-to-text configurations
│ │ │ ├── CreateEditProvider.tsx # Custom STT provider setup
│ │ │ ├── CustomProvider.tsx # Custom STT provider
│ │ │ ├── index.tsx # STT configs exports
│ │ │ └── Providers.tsx # STT providers list
│ │ └── SystemPrompt.tsx # Custom system prompt editor
│ ├── TextInput/ # Text input components
│ ├── ui/ # shadcn/ui component library
│ └── updater/ # Application update handling
├── config/ # Application configuration
│ ├── ai-providers.constants.ts # AI provider configurations
│ ├── constants.ts # General constants
│ ├── index.ts # Config exports
│ └── stt.constants.ts # Speech-to-text provider configs
├── contexts/ # React contexts
│ ├── app.context.tsx # Main application context
│ ├── index.ts # Context exports
│ └── theme.context.tsx # Theme management context
├── hooks/ # Custom React hooks
│ ├── index.ts # Hook exports
│ ├── useCompletion.ts # AI completion hook
│ ├── useCustomProvider.ts # Custom provider hook
│ ├── useCustomSttProviders.ts # Custom STT providers hook
│ ├── useSettings.ts # Settings management hook
│ ├── useVersion.ts # Version management hook
│ └── useWindow.ts # Window management hook
├── lib/ # Core utilities
│ ├── chat-history.ts # Chat history management
│ ├── functions/ # Core functionality modules
│ │ ├── ai-models.function.ts # AI model handling
│ │ ├── ai-response.function.ts # AI response processing
│ │ ├── common.function.ts # Common utilities
│ │ ├── index.ts # Function exports
│ │ └── stt.function.ts # Speech-to-text functions
│ ├── storage/ # Local storage management
│ │ ├── ai-providers.ts # AI provider storage
│ │ ├── helper.ts # Storage helpers
│ │ ├── index.ts # Storage exports
│ │ └── stt-providers.ts # STT provider storage
│ ├── index.ts # Lib exports
│ ├── utils.ts # Utility functions
│ └── version.ts # Version utilities
├── types/ # TypeScript type definitions
│ ├── ai-provider.type.ts # AI provider types
│ ├── completion.hook.ts # Completion hook types
│ ├── completion.ts # Completion types
│ ├── context.type.ts # Context types
│ ├── index.ts # Type exports
│ ├── settings.hook.ts # Settings hook types
│ ├── settings.ts # Settings types
│ └── stt.types.ts # Speech-to-text types
├── App.tsx # Main application component
├── main.tsx # Application entry point
├── global.css # Global styles
└── vite-env.d.ts # Vite environment types
src-tauri/
├── src/
│ ├── main.rs # Application entry point
│ ├── lib.rs # Core Tauri setup and IPC handlers
│ └── window.rs # Window management & positioning
├── build.rs # Build script for additional resources
├── Cargo.toml # Rust dependencies and build configuration
├── tauri.conf.json # Tauri configuration (windows, bundles, etc.)
├── Cargo.lock # Dependency lock file
├── capabilities/ # Permission configurations
│ └── default.json # Default capabilities
├── gen/ # Generated files
├── icons/ # Application icons (PNG, ICNS, ICO)
├── info.plist # macOS application info
├── pluely.desktop # Linux desktop file
└── target/ # Build output directory
# Start Tauri development
npm run tauri dev
# Build for production
npm run build
npm run tauri build
# Type checking
npm run type-check
# Linting
npm run lintWe welcome contributions! Here's how to get started:
💡 Like this project? Consider buying me a coffee ☕ or hiring me for your next project!
# Fork and clone the repository
git clone https://github.com/iamsrikanthnani/pluely.git
cd pluely
# Install dependencies
npm install
# Start development
npm run tauri dev- Fork the repository
- Create a feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
- TypeScript for type safety
- ESLint + Prettier for formatting
- Conventional Commits for commit messages
- Component documentation with JSDoc
This project is licensed under the MIT License - see the LICENSE file for details.
- Cluely - Inspiration for this open source alternative
- Tauri - Amazing desktop framework
- shadcn/ui - Beautiful UI components
- @ricky0123/vad-react - Voice Activity Detection
- OpenAI - GPT models and Whisper API
- Anthropic - Claude AI models
- xAI - Grok AI models
- Google - Gemini AI models
- 🎵 Audio Setup Guide: SYSTEM_AUDIO_SETUP.md - Complete cross-platform audio configuration
- Website: cluely.com (Original Cluely)
- Documentation: GitHub Wiki
- Issues: GitHub Issues
- Discussions: GitHub Discussions
Made with ❤️ by Srikanth Nani
Experience the power of Cluely, but with complete transparency and control over your data.