Scrape Pages with a Gemini Vision Agent and ScrapingBee

## Vision-Based Scraping AI Agent This is the central node of the workflow, powered by an AI Agent with two key prompts: - System Prompt: Instructs the AI on how and what data to extract from the screenshot. You can customize this to suit your needs. It also includes fallback instructions to call a tool for retrieving the HTML page if data extraction from the screenshot fails. - User Message: Provides the page URL for context. ### Sub-Nodes 1. Google Gemini Chat Model Chosen because tests show that Gemini-1.5-Pro outperforms GPT-4 and GPT-4-Vision in visual tasks. *Either my prompt wasn’t optimized for GPT models, or GPT might need glasses 👓*. Other multimodal LLMs haven’t been tested yet. 2. HTML-Based Scraping Tool A fallback tool the agent uses if it cannot extract data directly from the screenshot. 3. Structured Output Parser Formats the extracted data into an easy-to-use structure, ready to be added to the results page in Google Sheets.