TinyFish

TinyFish gives Claude a cloud browser and web agent to search, read, and operate the live web. Use it for web automation and browser automation on real websites: navigating pages, clicking through multi-step workflows, logging in to websites and account portals, filling out and submitting web forms, online applications, and job applications, and extracting structured data that only appears after real page interaction.

Three core tools cover the workflow. Web search finds relevant pages and news across the public web. Fetch reads a web page or webpage and extracts clean, token-efficient content, including JavaScript-rendered pages, so Claude can read, summarize, or scrape up to 10 URLs at once. The web agent automates website tasks that static fetching can't reach: it runs in a remote stealth Chrome browser with built-in anti-bot protection and proxy rotation, can reuse saved logged-in sessions to log in and check account status or dashboards, and returns results as structured JSON. The same task can run across many sites in parallel for price comparison and shopping: compare prices, check product availability and stock availability, gather ecommerce product data, product listings, job listings, and real estate listings.

Website monitoring: TinyFish monitors websites for changes. Set up a recurring monitor on a web page or search topic for price tracking, change detection, content updates, and alerts when something new appears. For direct control, open a headless cloud browser session with CDP access for Playwright, Puppeteer, or Selenium.

Common uses include web scraping and data extraction behind logins and navigation, a web scraper for tables, app reviews, and local businesses for lead generation, a social media scraper for public profiles and posts, website screenshots, public records and trademark search, downloading invoices, receipts, and billing statements from a customer portal, research across many sources, and pulling live data from web apps.