Scrapfly sells five distinct APIs and puts all of them behind a single key.
It also publishes operational figures that are unusually concrete: requests served in the billions, uptime at 99.99%, petabytes of data handled, tens of thousands of users.
The Five
The Web Scraping API for collection at scale. The Cloud Browser API giving remote Chromium sessions over CDP. The Crawler API for full site collection. The Data Extraction API returning structured fields. And the Screenshot API for full page captures.
Residential proxies, session handling and geo targeting sit underneath all of them.
The Cloud Browser Is the Standout
Managed Chromium controlled over the Chrome DevTools Protocol, which is a meaningfully different product from a scraping endpoint.
A scraping API gets you a page. A remote browser lets you drive one: click, scroll, fill forms, wait for things to appear, handle multi step flows.
Their own framing is the honest part. Headless browsers crash at scale, and anyone who has run a fleet of them knows exactly how true that is. Memory leaks, zombie processes, containers dying under load. Absorbing that operational misery is the actual product.
The Screenshot API Is More Useful Than It Sounds
Full page captures get dismissed as a novelty, and they are not.
Ad verification needs visual proof that an advert rendered. Compliance work needs evidence of what a page said on a given date. Visual regression testing needs images to compare. Disputes need something a human can look at.
Structured data cannot do any of that. Sometimes the picture is the deliverable.
The Crawler API Versus the Scraping API
The distinction matters and catches people out.
Fetching a known page is straightforward. Crawling a whole site is not: you have to discover URLs, respect structure, handle pagination, avoid loops and pace yourself so the target does not start refusing you.
That discovery logic is where home built crawlers usually break, and it is a genuinely different product from an endpoint that takes a URL. Having both means you can use the simpler one where it suffices.
Published Numbers Invite Comparison
Most providers quote pool sizes, which nobody can verify. Scrapfly quotes throughput, storage and uptime, which are operational rather than marketing claims.
Treat them as self reported, as always. But specificity of that kind is a reasonable signal, and a company inviting scale comparisons is usually one comfortable with the answer.
Pay for What You Use
Pricing starts at a modest monthly figure with per request costs published, scaling into enterprise tiers, and there are SDKs and integrations for common stacks.
One key across five products also means one balance, which avoids the usual problem of tracking several subscriptions for one project.
Who It Fits
Scrapfly suits work requiring real browser interaction rather than simple fetching, anyone needing screenshots as evidence, teams running full site crawls, and developers who want several capabilities without several vendors.
Who Might Look Elsewhere
If you only need IP addresses, a plain proxy provider is far cheaper.
Simple static targets do not need a cloud browser. Buy the cheapest thing that works.
Final Verdict
Scrapfly is one of the most complete platforms in this section and among the most technically credible.
Five APIs under one key, managed Chromium over CDP, full site crawling, structured extraction, screenshots, residential proxies underneath and published operational figures.
Start with whichever API matches your hardest problem rather than the cheapest one. Our scraping guide covers what you would otherwise be maintaining.
