ScraperAPI is built around a simple promise: send it a URL, and let the service handle the proxy and anti-bot infrastructure behind the request.
We tested that workflow hands-on using a new ScraperAPI trial account. Our goal was not to benchmark thousands of requests or stress-test the platform. We wanted to answer practical questions about the first API request, credit cost, JavaScript rendering, country-level routing, target errors, and whether the interface was understandable without building a scraping stack first.
The short version: ScraperAPI was easy to get running, basic requests were inexpensive in our test, and the cost difference for JavaScript rendering was very clear.
Affiliate disclosure: PG Systems Lab participates in the ScraperAPI affiliate program. We may earn a commission if you sign up through our affiliate link. This does not affect our testing or conclusions.
- What We Tested
- Signup and Trial
- First Impression of the Dashboard
- API Playground
- Standard HTML Request: 1 Credit
- What the Raw Response Looks Like
- JavaScript Rendering: 10 Credits
- Japan Geotargeting: Still 1 Credit
- Failure Test: Target 404 Was Preserved
- Cost Visibility Is One of the Better Parts
- ScraperAPI Credit Cost Planner
- Additional Options We Saw
- What We Liked
- What We Did Not Test
- Who Should Consider ScraperAPI?
- Who May Want Something Else?
- Our Verdict
- Testing Notes
What We Tested
- account signup and email activation
- ScraperAPI onboarding
- API Playground and generated Python request code
- a standard raw-HTML request
- JavaScript rendering
- Japan geotargeting with
country_code=jp - a deliberate 404 failure case
- request-credit estimates before execution
We did not test large-scale crawling, sustained concurrency, production proxy rotation at scale, paid-plan limits, residential or mobile proxy performance on heavily protected sites, or long-running DataPipeline jobs.
Signup and Trial
ScraperAPI offered a free trial with 5,000 API credits and no credit card required. Signup was available through Google, GitHub, or email; we used email signup. Email activation was required before access to the main dashboard.
The onboarding flow asked about our role and current data-collection method. We selected Founder / CxO / VP, In-house build, AI agent or LLM-based tool, and the use case “Scrape websites for raw data (HTML).”
First Impression of the Dashboard
The dashboard was more developer-focused than the email-marketing tools we have tested elsewhere. Navigation included Get started, API Playground, DataPipeline, Crawler, Analytics, Billing, MCP, additional APIs, and documentation.
The setup page exposed API, Structured Data, LLM-Ready Data, and AI Result Data access modes. For a developer or technical founder, the structure was fairly easy to understand.
API Playground
The API Playground was the most useful part of the first test. It let us build a request visually, choose options, and generate code in several languages. We used Python. The generated request was conceptually requests.get("https://api.scraperapi.com/", params=payload), with the API key and target URL supplied through request parameters.
The Playground showed generated code, copy-to-clipboard, code download, estimated cost per request, available credits, and optional DataPipeline scheduling. This made the first request easy to understand before execution.
Standard HTML Request: 1 Credit
For our first live request, we used https://quotes.toscrape.com/ with no advanced options enabled. ScraperAPI showed an estimated cost of 1 credit per request, and the request returned HTTP 200 OK.
The response contained the target page’s raw HTML, including the page title, quote text, author names, tags, and page structure. This was the simplest possible use case, and it worked as expected.
What the Raw Response Looks Like
The Playground displayed the full raw HTML rather than hiding it behind a simplified interface. Our response included normal HTML elements such as <title>, headings, links, quote blocks, metadata, and tags. That is useful for developers feeding the response into their own parser, application, or AI workflow.
JavaScript Rendering: 10 Credits
We then enabled JavaScript rendering. The Playground changed the request cost from 1 credit to 10 credits. The JavaScript-rendered request returned HTTP 200 OK.
The target page was mostly static, so this test does not prove that rendering exposed content that would otherwise have been unavailable. It did confirm that the rendering option worked, the request completed successfully, the cost difference was visible before execution, and JavaScript rendering cost 10× more credits than the basic request in this configuration.
Japan Geotargeting: Still 1 Credit
We selected country_code=jp and sent the same basic request. It returned HTTP 200 OK, and the Playground showed 1 credit per request.
The target page did not have meaningful location-specific content, so we are not claiming that the returned content changed based on location. We confirmed only that the parameter was accepted, the request succeeded, and the request remained priced at 1 credit in the tested configuration.
Failure Test: Target 404 Was Preserved
We intentionally requested https://quotes.toscrape.com/this-page-does-not-exist. The result was HTTP 404 Not Found, and the response body contained the target site’s normal 404 HTML.
ScraperAPI did not mask the target failure behind a generic successful API response in this test. For developers building retry logic, monitoring, or data-quality checks, preserving the target HTTP status can make error handling simpler.
Cost Visibility Is One of the Better Parts
- standard request: 1 credit
- Japan geotargeting: 1 credit
- JavaScript rendering: 10 credits
The dashboard also showed remaining credits. With 5,000 trial credits, the Playground made the equivalent basic-request capacity understandable under the tested conditions. This cost visibility helps users see which options become expensive before integrating them into production code.
ScraperAPI Credit Cost Planner
Estimate how many API credits a planned monthly workload may consume for common documented request modes before choosing a plan or enabling expensive options.
Estimate boundary: This is an estimate for common documented request modes, not a billing guarantee or an exact target-domain quote. Plan fit compares only with currently listed monthly credit allowances; it does not evaluate features, concurrency, geography, overages, custom terms, or price. Some domains, protected-domain behavior, account arrangements, and configurations consume different credits. Verify the current ScraperAPI Domain Cost Estimator, dashboard, documentation, and plan allowance for your exact workload.
The calculation runs only in your browser. It does not request an API key, target URL, email address, credentials, or billing-account data, and it does not send your calculator input to PG Systems Lab or an AI provider.
Affiliate disclosure: PG Systems Lab participates in the ScraperAPI affiliate program and may earn a commission if you sign up through this link.
This independent calculator is not an official ScraperAPI product and is not affiliated with or endorsed by ScraperAPI. Official sources: Credits and Requests Costs, Control and Optimization, and Pricing.
Additional Options We Saw
The API Playground exposed premium residential and mobile IPs, advanced bypass mechanisms, JavaScript rendering, country targeting, device type, max-cost controls, sessions, custom headers, and redirect behavior. Documentation described parameters such as premium=true, ultra_premium=true, render=true, country_code=..., device_type=mobile, and device_type=desktop.
We did not test all of these options, so their presence in the interface should not be read as a PG Systems Lab performance finding.
What We Liked
- easy first request
- clear API Playground
- generated code
- visible cost before execution
- raw HTML response
- simple Python integration
- geotargeting available through a request parameter
- JavaScript rendering available through a toggle
- target-side 404 behavior was visible
- no credit card required for the trial
The biggest positive was transparency: ScraperAPI made it easy to see when an option would make a request more expensive.
What We Did Not Test
- large request volumes or maximum concurrency
- long-running crawling
- difficult anti-bot targets or CAPTCHA-heavy sites
- residential/mobile IP success rates
- sustained JavaScript-rendering workloads
- paid-plan support
- DataPipeline or ScraperAPI Crawler at production scale
- MCP workflows
- structured-data accuracy
This review is a limited hands-on test, not a complete production benchmark. ScraperAPI does not guarantee bypass success or reliability on every website.
Who Should Consider ScraperAPI?
- developers who need raw web data
- technical founders
- teams building internal scraping workflows
- AI or LLM systems that need web inputs
- businesses that do not want to maintain proxy infrastructure
- teams that need country-level routing
- workflows where proxy and anti-bot handling would otherwise require additional engineering
Who May Want Something Else?
- you only need data from a few simple public pages
- direct HTTP requests already work reliably for your use case
- you need a fully managed research service rather than an API
- you need highly structured datasets instead of raw page access
- your use case does not justify extra credits for rendering or premium routing
Our Verdict
Our first ScraperAPI test was straightforward. The standard API request worked immediately, and the API Playground made the request understandable before execution. The generated code was simple enough to integrate into a normal Python workflow.
The clearest result was the cost comparison:
- Basic HTML request → 1 credit
- Japan geotargeting → 1 credit
- JavaScript rendering → 10 credits
We also confirmed that a deliberately invalid target returned a normal 404 response rather than being hidden behind a generic success response.
For teams that need to collect public web data without maintaining their own proxy and anti-bot infrastructure, ScraperAPI is worth testing against their own targets and workload.
Try ScraperAPI
Affiliate link. PG Systems Lab may earn a commission if you sign up through this link.
Testing Notes
This review reflects a limited hands-on test performed by PG Systems Lab in August 2026. Request pricing, product limits, feature availability, and trial terms may change. Always verify current ScraperAPI documentation and pricing before making production decisions.
