DeepSearch 🔍

DeepSearch is a modern web search application that combines the power of multiple AI models to provide comprehensive, well-cited search results. Built with Next.js and Tailwind CSS, it offers a seamless search experience with features like query refinement, smart summarization, and related searches.

✨ Features

• Smart Query Refinement: Automatically improves search queries for better results
• Multi-Provider Support:
  • Alibaba Cloud
  • DeepSeek
  • OpenAI
  • Tavily Search API
• Rich Search Results:
  • Comprehensive summaries with citations
  • Source links and snippets
  • Image previews when available
  • Related searches
• Modern UI/UX:
  • Dark/Light mode support
  • Responsive design
  • Real-time search suggestions
  • Parallelization for better performance

🚀 Getting Started

Prerequisites

• Node.js 18+ and npm (for local development)
• Docker and Docker Compose (for containerized deployment)
• API keys for:
  • Alibaba Cloud
  • DeepSeek
  • OpenAI
  • Tavily Search API

🖥️ Local Installation

1. Clone the repository:

   git clone https://github.com/yourusername/deep-search.git
   cd deep-search
   

2. Install dependencies:

   npm install
   

3. Create a .env.local file in the root directory:

   ALIBABACLOUD_API_KEY=your_alibabacloud_api_key
   DEEPSEEK_API_KEY=your_deepseek_api_key
   OPENAI_API_KEY=your_openai_api_key
   TAVILY_API_KEY=your_tavily_api_key
   

4. Start the development server:

   npm run dev
   

Visit http://localhost:3000 to start using DeepSearch.

🐋 Docker Installation (TODO)

1. Clone the repository and create .env.local as described above.

2. Build and run with Docker Compose:

   docker-compose up -d
   

This will:

• Build the Docker image with Node.js 18
• Set up the environment variables
• Start the container in detached mode
• Map port 3000 to your host machine

3. Access the application at http://localhost:3000

To stop the container:

docker-compose down

🔧 How It Works

1. Query Refinement:

   • User enters a search query
   • Selected AI provider (OpenAI/DeepSeek/Qwen) refines the query for better results
   • The refined query is used for the actual search

2. Search Process:

   • Tavily API performs the web search
   • Results include titles, snippets, URLs, and images

3. Summarization:

   • Selected AI provider analyzes search results
   • Generates a comprehensive research report with citations
   • Formats output in markdown with proper source attribution
   • Related searches are generated based on the report

4. Result Display:

   • Summary with clickable citations
   • Source list with links and snippets
   • Image previews from sources
   • Related search suggestions

🎨 UI Preview

TODO

🛠️ Configuration

API Providers

Configure API providers in .env.local:

ALIBABACLOUD_API_KEY=your_key
DEEPSEEK_API_KEY=your_key
OPENAI_API_KEY=your_key
TAVILY_API_KEY=your_key

Search Settings

Customize search behavior in src/lib/settings-context.tsx:

• Default provider
• Search depth
• Result count
• Image inclusion

📝 License

MIT License - feel free to use this project for your own purposes.

🚧 In Development

TODO