GPT4 Vision Web Crawling

This is a GPT4 Vision API and Puppeteer powered tool that will answer questions based on website screenshots. You ask it a question and it will browse to a website and take a screenshot. Then it will use GPT4 Vision API to answer the question based on the screenshot.

Quick Start

$ npm install
$ pip install -r requirements.txt
$ python3 vision_crawl.py

Examples

You can ask stuff like this, for example:

• "What is the weather like in California?"
• "What are the latest news in the world?"
• "What is the current stock price of Tesla?"