ScraperAPI
Two independent 2026 benchmarks converge on the same honest weak spots — 99-100% success on Amazon and GitHub, but Google drops to 82%, X/Twitter returns 0%, and Indeed and Zillow sit in the 51-60% range.
What is ScraperAPI?
ScraperAPI is a cloud-based scraping service built around a genuinely simple premise: send one API call with a target URL, and it handles proxy rotation across more than 40 million IPs, JavaScript rendering via built-in headless Chrome, CAPTCHA solving, and retry logic, returning already-unblocked, already-rendered page content — no need to run your own Playwright or Puppeteer instance, and no manual proxy management. It charges per successful request rather than by bandwidth, with four dedicated structured endpoints (/structured/amazon, /structured/ebay, /structured/google, /structured/walmart) that return parsed JSON directly rather than raw HTML you'd have to parse yourself. Genuinely honest user sentiment reflects a real strength in developer experience specifically: a 4.5/5 Trustpilot rating (93% five-star) and 4.4/5 on G2, with one reviewer citing 12 years of web data consulting experience describing the proxy rotation as "seamless" and crediting it with saving hours of debugging. Onboarding is fast, documentation is clear, and the API is genuinely forgiving of misconfiguration — real, meaningful advantages for teams that don't want to manage scraping infrastructure themselves.
The genuinely important, honest thing worth understanding clearly before choosing ScraperAPI for a specific target: two separate, independent 2026 benchmarks converge on the same real pattern of strengths and weaknesses. One test found 68.95% overall success (placing it in the lower performance tier for protected sites specifically), while performing far better on lightly protected domains — 99.21% on Amazon, 100% on GitHub — but dropping to 81.72% on Google, and returning zero results on X/Twitter entirely, since that platform's terms of service explicitly denylist scraping. A separate, later benchmark found 72.57% overall success at 5.6 seconds per request, again attributing the headline drop specifically to X/Twitter's 0% (the same ToS-based block) plus genuinely weak results on Indeed (60%) and Zillow (51%). Taken together, that's a fair, consistent picture: for teams scraping public data from sites without meaningful bot protection, ScraperAPI delivers strong, affordable, reliable results — but for Cloudflare-protected, DataDome-protected, or otherwise hardened targets specifically, that lower success rate translates directly into failed pipelines, and it's worth testing your actual target site before committing budget rather than assuming uniform performance across the whole web.
The two converging benchmark results are drawn directly from two separate, independent, dated 2026 sources (Bright Data's comparative guide, citing Proxyway's 2025 benchmark, and Scrape.do's own updated 2026 testing) — both reaching the same honest conclusion despite being run independently. Credit multiplier details are corroborated across multiple independent sources (Firecrawl, Proxies.sx). User sentiment and reputation figures are drawn from Firecrawl's aggregated review data.
Key features
Automatic proxy rotation
Draws from a pool of 40+ million IPs to prevent blocking, no manual management.
Built-in JavaScript rendering
Headless Chrome included, no need to run your own Playwright or Puppeteer instance.
Structured endpoints
Pre-parsed JSON directly from Amazon, eBay, Google, and Walmart.
CAPTCHA handling
Automatic bypass attempts to keep scraping running uninterrupted.
Geotargeting
Route requests through specific geographic locations for localized results.
Automatic retry logic
Failed requests are retried automatically without extra code.
Pricing
Free
- Genuinely usable for testing your actual target site first
- Test success rate on your specific site before committing budget
- Reported allowance varies slightly across sources
Hobby
- Basic requests cost 1 credit; JS rendering costs 5-10 credits
- E-commerce sites cost roughly 5x, search engines roughly 25x
- Genuinely affordable for under 100K pages/month on simple targets
Startup / Business
- Startup: 1,000,000 credits, 50 concurrent
- Business: 3,000,000 credits, 100 concurrent
- Enterprise custom pricing available for higher volume needs
Budget per target site, not on the blended average cost — pricing varies by an order of magnitude depending on how protected your specific target is. Prices reflect ScraperAPI's published pricing as of July 2026.
Available models
Integrations & platforms
Pros, cons & best for
Pros
- Genuinely simple, forgiving API with fast onboarding and clear documentation
- Strong, verified performance on lightly protected, standard targets like Amazon and GitHub
- Affordable pricing for teams under 100K pages/month on simple sites
Cons
- Genuinely weak, benchmarked performance on hardened, protected targets
- Credit costs multiply 5-25x depending on target complexity, hard to predict at first
- No mobile proxy IPs, and lower plans restrict to US/EU regions only
Best for
- Developers scraping public data from sites without heavy bot protection
- Teams wanting a simple, managed API without running their own proxy infrastructure
- Not the pick for Cloudflare or DataDome-protected targets specifically
Take a look inside
Alternatives
For a no-code interface or a larger actor marketplace instead:
Our verdict
ScraperAPI's genuine strength is real, developer-friendly simplicity — a single API call genuinely does replace proxy management, headless browser setup, and CAPTCHA handling, and its performance on lightly protected, standard targets like Amazon and GitHub is verified strong across independent testing. The honest, important thing worth taking seriously before committing to a specific target: two separate benchmarks independently converge on the same real weakness on hardened, heavily protected sites, with success rates dropping meaningfully on Google, Indeed, and Zillow specifically, and X/Twitter blocked entirely by policy rather than technical limitation. Credit costs that multiply 5-25x by target complexity are also worth budgeting for realistically rather than assuming a flat, predictable rate. For developers scraping public data from sites without heavy bot protection, ScraperAPI remains a genuinely strong, affordable, well-regarded choice; for Cloudflare- or DataDome-hardened targets specifically, test your actual site first before committing real budget.
FAQ
How reliable is ScraperAPI on protected websites?
Genuinely mixed, per two independent 2026 benchmarks — strong (99-100%) on lightly protected sites like Amazon and GitHub, but meaningfully weaker on hardened targets like Google (82%), Indeed (60%), and Zillow (51%), with X/Twitter returning 0% due to a policy-based denylist rather than a technical failure.
How much does ScraperAPI actually cost per request?
It varies significantly by target — a basic HTTP request costs 1 credit, JavaScript rendering costs 5-10 credits, e-commerce sites cost roughly 5x a basic request, and search engines cost roughly 25x, so budget per target site rather than on a blended average.
Do I need to manage my own proxies or browser automation with ScraperAPI?
No, that's the core value proposition — a single API call handles proxy rotation across 40+ million IPs and built-in headless Chrome rendering, so you don't need to run your own Playwright or Puppeteer setup.
Why does ScraperAPI return no results for X/Twitter?
Because X/Twitter's own terms of service explicitly denylist scraping of the platform, resulting in a 0% success rate that reflects policy enforcement rather than a technical scraping failure.
Is ScraperAPI's free tier good enough to test before committing?
Yes, it's genuinely useful for testing your specific target site's real success rate and cost before committing to a paid plan, given how much performance and pricing vary by target.
Does ScraperAPI support mobile proxy IPs?
No, and lower-tier plans are also restricted to US/EU regions only — this is a real, honest limitation worth knowing if your use case needs broader geographic or mobile proxy coverage.