Voice Virtual Assistant

Hands-free AI assistant with free (Ollama) and premium (OpenAI + ElevenLabs) options

[🎤 Demo Coming Soon]

A Python-based voice assistant with two implementations:

1. Free version - Uses Ollama (local AI) + macOS Text-to-Speech (completely free!)
2. Premium version - Uses ElevenLabs (realistic voices) + OpenAI GPT (paid APIs)

Features

• 🎤 Speech recognition using Google Speech Recognition
• 🤖 AI-powered conversational responses
• 🔊 Natural-sounding voice output
• 💬 Maintains conversation context
• 🎯 Simple voice commands
• 🆓 100% free option with local AI

Quick Start (FREE Version)

The free version runs completely locally with no API costs!

Prerequisites

• Python 3.8 or higher
• macOS (for built-in TTS)
• Microphone and speakers/headphones
• Ollama for local AI

Installation (Free Version)

1. Install Python dependencies:

   pip3 install SpeechRecognition requests

   Note for macOS users: If you encounter issues installing PyAudio:

   brew install portaudio
   pip3 install pyaudio

2. Install Ollama (local AI):

   brew install ollama

3. Download AI model:

   ollama pull llama2

4. Start Ollama service:

   brew services start ollama

Installation (Premium Version with APIs)

⚠️ WARNING: This version requires paid API services

1. Install all dependencies:

   pip3 install -r requirements.txt

2. Set up environment variables:

   cp .env.example .env

   Edit .env and add your API keys:

   ELEVENLABS_API_KEY=your_elevenlabs_api_key_here
   OPENAI_API_KEY=your_openai_api_key_here
   ELEVENLABS_VOICE_ID=21m00Tcm4TlvDq8ikWAM
   

   Get API keys:

   • ElevenLabs: https://elevenlabs.io/ (requires payment)
   • OpenAI: https://platform.openai.com/ (requires payment)

Usage

Free Version (Recommended)

Run the free assistant:

python3 assistant_free.py

Premium Version

Run the premium assistant (requires API keys):

python3 assistant.py

Voice Commands

• Speak naturally to ask questions or have conversations
• Say "exit", "quit", or "goodbye" to stop the assistant
• Say "list voices" to see available ElevenLabs voices (premium only)
• Press Ctrl+C to force exit

Customization

Free Version

Change AI Model: Edit assistant_free.py line 68 to use a different model:

"model": "llama3.2",  # Options: llama2, llama3.2, mistral, codellama, etc.

Install new models with:

ollama pull llama3.2
ollama pull mistral

Adjust Voice: The free version uses macOS say command. Change the voice:

say -v "?"  # List available voices

Edit line 97 in assistant_free.py:

subprocess.run(["say", "-v", "Samantha", text], check=True)

Premium Version

Change Voice:

1. Run the assistant and say "list voices" to see available options
2. Copy the desired voice ID
3. Update ELEVENLABS_VOICE_ID in your .env file

Adjust AI Model: Edit assistant.py line 71 to change the model:

model="gpt-4o-mini",  # Options: gpt-4o-mini, gpt-4o, gpt-3.5-turbo, etc.

Modify System Prompt: Edit the conversation_history initialization in assistant.py to change the assistant's personality:

conversation_history = [
    {"role": "system", "content": "Your custom prompt here"}
]

Troubleshooting

Free Version

"Ollama is not installed or not running":

brew services start ollama
# Wait a few seconds, then try again

"API returned status code: 404": The AI model isn't downloaded:

ollama pull llama2

Slow AI responses:

• First response is always slower (model loading)
• Use a smaller/faster model: ollama pull llama3.2:1b
• Upgrade your Mac's RAM or use a smaller model

Both Versions

Microphone Issues:

• Ensure your microphone is connected and selected as the default input device
• Grant microphone permissions: System Settings → Privacy & Security → Microphone → Enable for Terminal
• Speak closer to the microphone
• Reduce background noise

"Could not understand audio":

• Speak more clearly and slowly
• Check that your microphone is working: say "test"
• Try adjusting energy_threshold in the code (line 39)

Audio Playback Issues:

• Ensure speakers/headphones are connected
• Check system audio settings
• Test with: say "test"

Premium Version Only

API Errors:

• Verify your API keys are correct and have available credits
• Add payment method to OpenAI and ElevenLabs accounts
• Check your internet connection

Cost Considerations

Free Version

• ✅ $0 - Completely free!
• Uses local AI (Ollama)
• Uses macOS built-in TTS
• Google Speech Recognition (free tier)

Premium Version

• 💰 OpenAI: ~$0.01-0.03 per conversation (GPT-4o-mini)
• 💰 ElevenLabs: ~$0.01-0.05 per response (depends on length)
• ✅ Google Speech Recognition: Free for basic usage

Estimated cost per 10-minute conversation: $0.50-$2.00

Future Improvements

• Add wake word detection ("Hey Assistant")
• Implement conversation memory across sessions
• Add web UI for configuration
• Support for other languages
• Integration with smart home devices
• Better formant preservation for voice responses
• Context-aware responses with RAG (Retrieval Augmented Generation)

Sample Conversations

Example 1: General Knowledge

You: What is the capital of France?
Assistant: The capital of France is Paris, which is also the country's largest city...

You: What's the population?
Assistant: Paris has a population of approximately 2.2 million within the city limits...

Example 2: Task Assistance

You: Remind me to call John at 3 PM
Assistant: I've set a reminder for you to call John at 3 PM today.

You: What's the weather like?
Assistant: I can help with that! However, weather integration isn't enabled yet...

Example 3: Conversational Context

You: My name is Sarah
Assistant: Nice to meet you, Sarah! How can I help you today?

You: What's my name?
Assistant: Your name is Sarah.

Use Cases

• 💻 Hands-free computing - Control your computer while multitasking
• 🎯 Learning assistant - Ask questions while studying
• ♿ Accessibility - Voice interface for those who need it
• 🏠 Smart home - Voice control (with future integrations)
• 📝 Note-taking - Dictate notes and ideas
• 👨‍💻 Productivity - Quick answers without leaving your workflow
• 🎓 Language learning - Practice conversation skills

Comparison: Free vs Premium

Feature	Free Version	Premium Version
Cost	$0	~$0.50-$2/conversation
AI Quality	Good (Llama2)	Excellent (GPT-4o-mini)
Voice Quality	Natural (macOS TTS)	Highly realistic (ElevenLabs)
Response Speed	2-5 seconds	1-3 seconds
Privacy	100% local	Data sent to APIs
Setup Complexity	Easy	Medium (API keys)
Internet Required	No (after model download)	Yes

License

MIT License - Feel free to modify and use for your projects!