ScraperAPI
Simple pay-per-credit scraping API
The simplest scraping endpoint to wire up, and the one most likely to surprise you on the invoice when a target needs rendering or fights back.
Developers scraping easy and medium sites at volume who want proxies, CAPTCHAs, and rendering handled in one GET request.
Credits are not requests. Per-domain and feature multipliers stack non-linearly, so 100K credits can shrink to a few thousand real pulls.
Reach for it when your targets are open or lightly defended and you want minimum setup. Skip it for social media or hard anti-bot sites.
What you're actually getting
ScraperAPI is a single HTTP endpoint you send a URL to. It handles proxy rotation across 40M-plus IPs in 50-plus countries, CAPTCHA solving, JavaScript rendering, and automatic retries, then hands back HTML or parsed JSON.
The draw is structured data endpoints for Amazon, Google, Walmart, and similar high-demand domains. You get clean JSON with product titles, prices, and ratings instead of raw HTML to parse yourself.
Around the core API sit an Async Scraper for millions of queued jobs, a no-code DataPipeline scheduler, and an MCP server that lets Claude and LangChain agents pull live web data. Founded in 2018 and now owned by SaaS.group, it claims 10,000-plus brands and 36 billion requests a month.
It's impressive that this API supports LLM-ready data, which makes it much easier for me to use with language models and quickly retrieve information. Additionally, the analytics provide a clear overview of the system's status, and the API playground simplifies testing requests compared to working in a separate environment. Everything is well done overall, but the pricing seems a bit off to me. I wish the credits were a bit more affordable.
Where it earns its keep
It shines on open and lightly defended sites. Independent benchmarks put Zillow at 97 percent success, Amazon at 92, Walmart at 87, and LinkedIn at 87.
The structured endpoints are the real time-saver. A Google SERP or an Amazon product page comes back as clean JSON, no selector wrangling and no parser maintenance. For a GTM team that means competitor price monitoring or search-rank tracking without hiring anyone to maintain scrapers.
For a developer who wants proxies and unblocking as a service, the integration is genuinely minutes. One API key, one base URL, flip a render or country parameter. Review scores sit at 4.4 on G2 and 4.6 on Capterra.
Where it'll bite you
Credits are not requests, and that is the trap. A normal page costs 1 credit, Amazon 5, Google SERP 25, LinkedIn 30. Rendering adds 10, premium proxies 10, ultra-premium 30, and the stacking is non-linear: ultra-premium plus rendering costs 75 credits, not 40.
That means a 100K-credit Hobby plan can be as few as 1,333 real pulls on a hard, JavaScript-heavy target. Credits do not roll over, and reviewers report being billed for CAPTCHA or error pages returned as successful.
Hard anti-bot sites are the second wall. The same benchmark scored Instagram, Twitter, and Booking at 0 percent success. If your target list leans social or heavily defended, ScraperAPI is the wrong tool.
I got serious problems trying to scrape reviews from Amazon, I'm using ScraperAPI but it keeps blocking me - any suggestion?
What it costs, really
There is a standing free tier of 1,000 credits a month plus a 7-day trial of 5,000 credits, no card required. Paid runs from $49/mo for 100K credits (Hobby) through $149 (1M), $299 (3M), $475 (5M), $975 (10.5M), to $1,975/mo for 21.5M (Advanced), with Enterprise custom.
The sticker price barely matters. The credit multiplier decides your real cost, so price your own URL mix in the dashboard's Domain Cost Estimator before you commit to any tier.
Same job, other tools
Also filed under Web data & extraction infrastructure:
