Ultrascrape

Scrape Reddit, Amazon, Walmart, YouTube, Trustpilot and any page with your own browser, for Claude, Codex or CSV/JSON export.

Ultrascrape collects customer comments, reviews and web page content from Reddit, Amazon, Walmart, YouTube, Trustpilot and other websites, then exports the results to CSV or JSON or connects them to your AI assistant. HOW IT WORKS • Open a supported thread, product page, video or web page, click Ultrascrape, and export the available content as CSV or JSON. • Connect the Ultrascrape MCP server to Claude Code, Claude Desktop or Codex and ask for research in plain English. For example: “What do people dislike about standing desks? Use Reddit and Amazon reviews.” • Your AI assistant can analyze the results to find recurring objections, praise and customer insights. SUPPORTED SOURCES • Reddit: search, subreddits and comment threads • Amazon: search, product details and reviews by star rating; some reviews require you to sign in to Amazon • Walmart: search, products and reviews • YouTube: search and comments • Trustpilot: company reviews, filterable by stars • Any supported web page: readable text, links and structured data PRIVACY Scraping runs in your own Chrome browser. Results are saved to files you choose or passed to your locally connected AI tool. Ultrascrape's servers do not receive scraped content. If you use an AI provider, its own data policies apply. The extension's CSV/JSON export works without the MCP server. AI integration requires the separate Ultrascrape MCP server. Website access, sign-in requirements and available content can limit extraction. MCP plans include a free allowance of 25 scrapes per month. Optional paid plans are available. Setup instructions, plan details and support: https://ultrascrape.vercel.app