Core product types / features
- API – one API to turn the web into structured, LLM‑ready data
- Search (web search + page content)
- Scrape (URL → clean markdown / JSON / text / screenshots)
- Interact (cloud sandboxes / agents acting on pages)
- Crawl (crawl entire websites)
- Map (map all URLs / site structure)
- Structured data (LLM‑ready)
- Clean markdown
- Structured JSON
- Screenshots
- Semantic text
- Page content (full page, not just links)
- Real‑time context / fresh knowledge
System / infra‑style types
- Agent (AI agent, “intelligence as core”)
- RAG pipeline (retrieval‑augmented generation)
- AI‑native software (foundation for)
- MCP (Agent tools with MCP / “live web access with MCP”)
- Cloud sandbox (for agents)
- Endpoint
- Onboarding (how teams get started)
Use‑case / application types
- Deep research
- Multi‑step web research (with live data)
- Smarter AI chats
- AI assistants
- AI agent tools
- Lead enrichment
- Sales pipeline building
- Web data (in general, “wherever it lives”)
Stages / modes
- Research Preview
Here’s a summary of the software items mentioned, grouped by category:
Headless Browser APIs
- Splash, Zombie.js, SimpleBrowser, DotNetBrowser
Browser Automation (Unified Interface)
- Selenium WebDriver, Playwright, Puppeteer
Test Automation
- Capybara, Jasmine, Cypress, QF-Test
Browser API Alternatives
- Deno (built-in browser APIs), jsdom (for Node.js), HtmlUnit (Java-based, uses Rhino engine for JS/Ajax)