LUMA: Your Personal "Jarvis"
Intelligent, responsive, and always ready to assist.
Discover Features
Why LUMA?
Seamless Voice-to-Voice
Fluid conversations with real-time speech recognition and natural text-to-speech output.
Autonomous Web Agent
Integrated AGNO agent with Gemini API for local web surfing and task management.
AI Powered
Driven by Groq LLaMA 3.1 and Google Gemini for high-speed, intelligent responses.
Beautiful Terminal UI
A clean, interactive interface for your terminal, complete with session statistics.
Core Capabilities
- Seamless voice-to-voice conversations
- Real-time speech recognition (using Moonshine ASR)
- Natural language AI responses (using Groq LLaMA 3.1)
- Text-to-speech output (using pyttsx3)
- Beautiful terminal interface
Smart Features
- Context-aware conversations
- Automatic speech detection (using Silero VAD)
- Real-time processing with audio buffering
- Command system for fine-grained control
- Session statistics tracking
🛠️ How It Works (Architecture)
LUMA's functionality is divided into three core component groups:
Audio Processing:
- Raw audio input capture
- Silero VAD for speech detection
- Audio buffering and chunking
Speech Processing Pipeline:
- Moonshine ASR for speech-to-text
- Groq LLaMA 3.1 for language processing
- pyttsx3 for text-to-speech synthesis
System Components:
- Terminal UI for interaction
- Session statistics tracking
- Command handler system
Get Started in Minutes
Step 2: Run the Installer
irm https://raw.githubusercontent.com/mc095/LUMA/main/install.ps1 | iex
curl -fsSL https://raw.githubusercontent.com/mc095/LUMA/main/install.sh | bash
The installation script will:
- Create a new directory for LUMA in a safe location.
- Download the latest secure LUMA executable.
- Prompt you to enter your API keys.
- Create your local configuration file automatically.
Your Security is Our Priority
LUMA runs 100% locally. Your API keys are never sent to any server. They are stored securely in a configuration file *only on your computer*.
💡 Usage Tips
- Ensure your microphone is connected and working.
- Speak clearly and naturally.
- For current events, include words like "latest" or "what's new on...".
- LUMA will respond both verbally and with on-screen text.
- Say "goodbye" or press Ctrl+C to exit the application.
🔧 Manual Installation (Advanced)
Show
Hide
If you prefer to install manually:
1. Download LUMA executable:
curl -L -o LUMA.exe https://github.com/mc095/LUMA/releases/latest/download/LUMA.exe
2. Create build_config.py with your API keys:
echo """Build configuration with embedded API keys.""" > build_config.py
echo GROQ_API_KEY = "your-groq-key-here" >> build_config.py
echo GEMINI_API_KEY = "your-gemini-key-here" >> build_config.py