This project provides a streamlined interface to generate concise, markdown-formatted summaries of company landing pages. It leverages large language models (LLMs) to extract and articulate key information suitable for potential customers, investors, or recruits.
The application allows users to:
Input a company name and URL
Choose from three LLMs: Qwen, DeepSeek, or LLaMA
Automatically scrape and clean website content
Generate a structured summary in real-time
Interactive user interface built with Gradio
Web scraping via BeautifulSoup
Real-time streaming of model output
Support for multiple models via OpenRouter
Python
Gradio (UI framework)
BeautifulSoup (HTML parsing and cleaning)
OpenRouter API (multi-model access)
dotenv (secure environment variable management)
git clone https://github.com/Islam-Reda-13/llm-website-summary.git
pip install -r requirements.txt
OPENAI_API_KEY=your_key_here
OPENROUTER_API_KEY=your_openrouter_key_here
python main.py
├── main.py # Gradio interface and app layout
├── models_and_prompting.py # Prompt handling and model streaming logic
├── web_scraping.py # HTML content scraper using BeautifulSoup
└── requirements.txt # Python package dependencies```