Documentation

Installation

1. Download and install Ollama
2. Clone the Companion repository
3. Install Python dependencies: pip install -r requirements.txt
4. Run the application: python main.py

First Chat

Once Companion is running:

• Select a model from the dropdown
• Type your message in the chat input
• Press Enter or click Send
• View the AI response in the chat window

Local Models (Ollama)

Companion automatically detects Ollama models. To add more models:

ollama pull <model-name>

Cloud Models

Cloud AI models are available in Web Mode and provide access to the latest AI capabilities including GPT-4, Gemini, Claude, and more.

Keyboard Shortcuts

• Enter - Send message
• Ctrl+Enter - New line in message
• Ctrl+K - Clear chat
• Ctrl+S - Save conversation

Chat Management

• Conversations are automatically saved
• Use the Clear button to start fresh
• Export conversations in multiple formats

Web Mode vs Desktop Mode

Companion offers two ways to use AI:

🌐 Web Mode (Cloud AI Only)

• Use directly in browser - no download needed
• Access to cloud AI models (GPT-4, Gemini, Claude)
• Always up-to-date with latest features
• Requires internet connection
• Perfect for quick testing and mobile use

💻 Desktop Mode (Cloud + Local AI)

• Download and install full application
• Access to both cloud AND local AI models
• Works offline with local models
• Enhanced privacy and performance
• Best for power users and regular use

Choosing the Right Mode

Choose Web Mode if you:

• Want to try Companion quickly
• Primarily use cloud AI models
• Use multiple devices including mobile
• Don't want to install software

Choose Desktop Mode if you:

• Want offline AI capabilities
• Value privacy and data control
• Use AI regularly for work
• Want to minimize cloud costs

📖 Read the full AI Modes documentation

Common Issues

• Ollama not found: Make sure Ollama is installed and running
• No models available: Pull models using ollama pull
• Cloud models not working: Check your internet connection
• Connection issues: Verify firewall settings

Getting Help

Need more help? Check out our:

• GitHub Repository
• Issue Tracker
• Community Forum

Integrate Companion's AI capabilities into your applications with our comprehensive REST API.

🔌 RESTful API

Simple HTTP endpoints for conversation management and AI interactions

🤖 AI Models

Access to multiple advanced AI models with intelligent fallback:

• GPT-4o - OpenAI's most advanced model
• Claude 3.5 - Anthropic's reasoning powerhouse
• DeepSeek R1 - Advanced reasoning capabilities
• Perplexity - Web-connected research model
• Gemini 2.5 Flash - Google's fast multimodal AI

🌟 Enhanced Capabilities

Built-in features for advanced AI interactions:

• 🌐 Web Search - Real-time information from multiple search engines
• 🧠 NLP Analysis - Natural language processing and understanding
• 📈 Learning - Adaptive responses based on context
• ⚡ Intelligent Caching - Time-sensitive data management

📝 Quick Start Examples

Create & Manage Conversations

// Create a new conversation
POST /api/conversations
Content-Type: application/json

// Response
{
  "id": "uuid-string",
  "title": "New Chat",
  "created_at": "2024-08-14T18:30:00Z"
}

// Get all conversations
GET /api/conversations

// Delete a conversation
DELETE /api/conversations/{id}

Send Messages & Get AI Responses

// Send a message
POST /api/conversations/{id}/messages
Content-Type: application/json

{
  "message": "Explain quantum computing",
  "active_tools": ["web", "think"]
}

// Response with AI answer
{
  "content": "Quantum computing is...",
  "thinking": "Step-by-step reasoning...",
  "web_sources": ["url1", "url2"]
}

AI Tools & Capabilities

// Available AI tools
{
  "web": "Real-time web search",
  "think": "Step-by-step reasoning", 
  "deepsearch": "Comprehensive research",
  "code": "Programming assistance",
  "deepthink": "Multi-model analysis"
}

// Health check
GET /api/health

// System statistics  
GET /api/stats

🔗 API Actions