Migration from ScrapingBee

If you already have jobs built around ScrapingBee, the Toolkit API scrape endpoint should feel familiar.

Need Python snippets instead of raw request mappings? See the dedicated guide: Python SDK examples for ScrapingBee users.

Parameter mapping

ScrapingBee	Toolkit API
`url`	`url`
`render_js=true`	`render_js: true`
`return_page_markdown=true`	`output: "markdown"`
`return_page_text=true`	`output: "text"`
`extract_rules`	`extract.selectors`
`ai_extract_rules`	`ai_extract.schema`
`wait_for`	`wait_for`
`wait_browser`	`wait_until`
`block_resources`	`block_resources`
`premium_proxy` / `country_code`	`proxy`
cookies / custom headers	`cookies` and `headers`
session reuse	`session_name`

Example translation

ScrapingBee-style idea

{
  "url": "https://toolkitapi.io/product/123",
  "render_js": true,
  "wait_for": ".price",
  "block_resources": true,
  "extract_rules": {
    "title": "h1",
    "price": ".price"
  }
}

Toolkit API equivalent

{
  "url": "https://toolkitapi.io/product/123",
  "render_js": true,
  "wait_for": ".price",
  "block_resources": ["image", "font"],
  "output": "clean",
  "extract": {
    "selectors": {
      "title": "h1",
      "price": ".price"
    }
  }
}

What you gain

Cleaner built-in Markdown and text modes for AI and search workflows
Richer structured sections in a single response instead of stitching together multiple endpoints
Dedicated SEO endpoints for audit, comparison, readability, and performance checks
Built-in async crawling support when you need to move beyond one-page extraction
Simpler LLM/RAG workflows through chunking and schema-driven ai_extract

Practical difference in developer experience

With many scraping services, the first successful fetch is only the beginning — you still need extra parsing, selector logic, and follow-up metadata calls. Toolkit API is designed to reduce that glue code.

Instead of thinking in terms of “fetch HTML, then parse, then enrich,” you can think in terms of “describe the content and signals I need, and get them back in one structured response.”

Recommended migration strategy

Move your simplest HTML or Markdown jobs first
Replace selector extraction with extract.selectors
Add JS rendering only where plain fetch is not enough
Adopt the SEO and crawl endpoints for workflows that previously required extra tooling
Standardize on markdown or clean output for AI-facing pipelines