An advanced AI-powered deep research application that synthesizes information from numerous sources to generate comprehensive, well-documented reports on complex topics.
KResearch is a sophisticated research assistant designed to tackle complex topics by leveraging a multi-agent AI system. It automates the process of in-depth research by planning, executing, and synthesizing information from the web. The final output is a comprehensive, well-structured report complete with source citations and a visual knowledge graph, making it an invaluable tool for students, analysts, and anyone needing to quickly develop a deep understanding of a subject.
- Conversational AI Agents: Utilizes 'Alpha' (Strategist) and 'Beta' (Tactician) agents who collaborate to create an optimal research plan.
- Iterative Research Cycles: Performs multiple cycles of planning, searching, and reading to gather comprehensive insights.
- Real-time Progress Tracking: Visualizes the AI's entire thought process in a detailed, step-by-step timeline.
- Configurable Research Modes: Offers 'Balanced', 'Deep Dive', 'Fast', and 'Ultra Fast' modes to tailor the research process to your needs.
- Comprehensive Final Report: Generates a well-structured final report in Markdown, synthesizing all findings.
- Knowledge Graph Visualization: Automatically creates a Mermaid.js graph to visualize key entities and their relationships.
- Sourced Citations: Grounds all research using Google Search and provides a complete list of sources.
- Modern & Responsive UI: A sleek, glassmorphism design built with React and Tailwind CSS, featuring light and dark modes.
This section will guide you through setting up and running the KResearch application locally.
You must have a Google Gemini API key to use this application.
- Google Gemini API Key: Obtain from Google AI Studio.
This project is a static web application but uses a local development server for better module handling and live reloading.
- Clone the repository:
git clone https://github.com/KuekHaoYang/KResearch cd research - Install dependencies:
Make sure you have Node.js installed, then run:
npm install
- Start the development server:
npm run dev
- Open in browser:
Once the server is running, open your browser and navigate to the local address provided (e.g.,
http://localhost:5173,http://localhost:3000, etc., depending on your setup).
The application requires the Google Gemini API key to be available as an environment variable.
Create a .env file in the root of the project (if your serving method supports it) or ensure the API_KEY is set in the environment where the application is hosted.
# .env file
API_KEY="YOUR_GEMINI_API_KEY"The application code directly accesses this key via process.env.API_KEY. You must ensure your local server or deployment environment makes this variable accessible to the frontend code.
Once the application is running in your browser:
- Select a Research Mode: Choose from 'Balanced', 'Deep Dive', 'Fast', or 'Ultra Fast' depending on the desired depth and speed.
- Enter Your Query: Type your research topic or question into the main text area.
- Start Research: Click the "Start Research" button or press
Enter(withoutShift). - Monitor Progress: Observe the research log as the AI agents work. You can stop the process at any time.
- Review Results: Once complete, the final report, knowledge graph, and citations will be displayed. You can copy the report text to your clipboard.
- New Research: Click "Start New Research" to clear the results and begin again.
Contributions are what make the open-source community such an amazing place to learn, inspire, and create. Any contributions you make are greatly appreciated.
Distributed under the MIT License. See LICENSE for more information.
- Powered by the Google Gemini API.
- UI inspired by modern glassmorphism design trends.