Scraperapi

ScraperAPI returns the scraped page's status code in the X-StatusCode response header, plus its own HTTP status: 200 for successful scrapes (even if the target page returned a 404), 500/503 for ScraperAPI-side failures. Failed JavaScript renders, CAPTCHA failures, and IP blocks return appropriate status codes. The async endpoint returns a job ID for polling. Retry logic for blocked requests is handled internally. Error responses include a message field. Credit depletion returns 401.

Authentication via api_key query parameter or X-ScraperAPI-Key header. Account-level API key — no per-domain or per-request scoping. Key rotation is manual via the dashboard. Usage is tracked by credits (each request costs 1-100 credits depending on options). No OAuth. The proxy pattern requires adding the API key to the proxy URL. Standard API key model with credit-based metering.

Documentation at scraperapi.com/documentation covers the API reference, proxy usage, structured data extraction, async mode, and code examples in Python, Node, PHP, Ruby, Java, and Go. The documentation is well-organized and practical. Getting started guide is concise. A Postman collection is available. Community support via chat and email. Documentation quality is good; CAPTCHA success rates and geo-coverage limitations are not quantified.

Two API patterns: (1) Direct HTTP API: GET https://api.scraperapi.com?api_key=KEY&url=TARGET[&options] — returns HTML directly; (2) Proxy endpoint: set ScraperAPI as an HTTP proxy, requests pass through transparently. Both patterns support the same option set: render=true (JS rendering), country_code, premium (residential proxies), session_number (sticky session), autoparse (structured extraction for Amazon/Google). The structured data endpoint (GET /structured) returns JSON for supported site types. API response is the raw HTML or JSON for structured mode.

ScraperAPI is a web scraping proxy service that handles proxy rotation, CAPTCHA solving, and JavaScript rendering so agents can focus on extracting data rather than infrastructure. The API is simple: an agent sends a request through ScraperAPI with a target URL, and receives the rendered HTML. This is directly useful for agents that need to extract content from arbitrary web pages without managing browser infrastructure. Options include JavaScript rendering, geolocation, structured data extraction for specific site types, and async mode for long-running scrapes. Free tier: 1,000 requests/month. Confidence is docs-derived; CAPTCHA success rates and rendering reliability are not runtime-verified.