ScraperAPI Pricing and Real Cost per 1,000 Pages by Target Site

Every scraping API publishes a monthly price and a credit allowance, and the two numbers together still do not tell you what you will pay. The variable that decides your bill is how many credits each page costs, and that depends entirely on which sites you point the crawler at. A team scraping small business websites and a team scraping search results can sit on the same plan and get bills that differ by a factor of twenty five.

This piece converts ScraperAPI's published pricing into a single comparable number, dollars per 1,000 pages, for the targets people actually scrape. Prices below were read from ScraperAPI's own pricing page and credit documentation on 20 September 2026. Vendors change them, so check before you sign anything.

The plans

ScraperAPI sells monthly credit bundles with a concurrency limit attached. Annual billing takes 10 percent off, and new accounts get a 7 day trial with 5,000 credits and no card required.

Notice the jump between Hobby and Startup. Tripling the price multiplies the allowance by ten, which means anyone hitting their Hobby ceiling every month is paying roughly three times more per credit than they need to. That step is the single biggest saving available on this pricing page, and plenty of teams never look at it.

What a credit costs

Divide price by credits and the plans line up like this: Hobby works out at 0.49 USD per 1,000 credits, Startup at 0.15, Business at 0.10, Scaling at 0.095, and the two largest plans at just over 0.09. Past Business the curve is basically flat, so volume discounts stop mattering and concurrency becomes the thing you are buying.

The multipliers

Here is where the real money is. ScraperAPI charges a different number of credits depending on the domain category and the parameters you set.

Two of those deserve a flag. Rendering and the anti-bot bypass are both add-ons, so a Cloudflare-protected page that also needs a browser is 21 credits rather than 1. And the SERP category covers every Google and Bing subdomain, which quietly includes Google Maps and Google Shopping, not only the classic results page.

To their credit, ScraperAPI only bills successful responses, defined as status 200 and 404, plus requests you cancel before their 70 second window is up. That is the right way to do it and it is worth confirming with any vendor you compare, because paying for blocks turns a hard target into an open-ended bill.

Cost per 1,000 pages

Now the useful table. This is what 1,000 pages of each type costs on the 49 USD Hobby plan and on the 299 USD Business plan, next to flat per request pricing at a comparable spend. Our own plans are Starter at 49 USD for 50,000 successful requests, which is 0.98 USD per 1,000, and Scale at 499 USD for 1,500,000, which is 0.33 USD per 1,000, with no multiplier for a difficult target.

Read the first row before anything else, because it is the one that goes against us. For plain, undefended HTML, credit pricing is cheaper, twice as cheap at the entry tier and three times cheaper at Business. If your crawl is mostly small sites, documentation, directories or news pages that return their content in the first response, credits are the better deal and you should buy them.

Every row below the first one moves the other way, and it moves hard. A daily rank tracking job is the clearest case: 1,000 queries a day is 30,000 results pages a month, which at 25 credits each is 750,000 credits, so you need the 149 USD Startup plan. The same 30,000 requests fit a 49 USD flat plan with 20,000 to spare. That gap is not a rounding error, it is the difference between a project that gets approved and one that does not.

The breakeven

There is a clean rule hiding in the arithmetic. Work out your average credits per page, weighted across the targets you actually crawl, and compare it to the breakeven for the tier you are on. At the 49 USD level the two models cost the same at 2 credits per page. At the 300 to 500 USD level they meet at about 3.3 credits per page. Below that number, credit pricing is cheaper. Above it, flat per request pricing is, and the gap widens with every multiplier you trigger.

Most real crawls land above the line without anyone noticing, because the easy pages get scraped once and the hard pages get scraped every day. Your average is not the average of your URL list, it is the average of your request volume, and the defended targets are the ones you re-check.

Where credit pricing is the right buy

• Your targets are mostly undefended pages that return content in the raw HTML.
• You want one vendor with prebuilt structured endpoints for a long list of sites and you are happy to pay the category rate for the convenience.
• Volume is very large and uniform, where the per credit price past the Business plan is genuinely low.
• You need concurrency in the hundreds, since that is what the upper plans are really selling.

Where flat per request pricing is the right buy

• Search results. One request returns one results page whether it is Google, Bing or anything else, which is how the SERP API is billed here.
• Marketplaces. Product, price, seller and availability come back at the same unit cost as any other page through the Amazon scraping API.
• Anything behind Cloudflare, DataDome or PerimeterX, where the bypass is the normal case rather than an exception. That is the whole premise of the Cloudflare scraping API and the DataDome bypass API.
• Mixed crawls, where you cannot predict next quarter's target list and do not want your budget re-priced when a site turns on bot management.
• Anything that needs a rendered browser on most requests, since rendering is included rather than charged at 10 credits a call.

The honest summary is that these are two different bets. Credit pricing bets that your pages are easy and charges you when they are not. Flat pricing bets that some of your pages are hard and charges a little more for the easy ones to average it out. Neither is a trick, and picking the wrong one is what costs money.

Estimate your own bill in ten minutes

You do not need a spreadsheet model for this. Four steps:

• Count by target, not by URL. List your sources and write down how many requests a month each one takes, including the re-checks. A single competitor checked hourly is 720 requests a month, not one.
• Assign credits to each line. Use the multiplier table above. Be pessimistic about rendering, because the pages that need it are rarely the ones you expect.
• Add a retry allowance. Defended targets fail on the first attempt often enough to matter. With any vendor that bills only successes this is free, but it does affect the throughput you need.
• Divide, then compare. Total credits times the per credit rate of the plan you would actually sit on, against total requests times the flat rate. Then look at the number, not at the plan name.

That exercise is the same discipline finance already applies elsewhere, and the teams who keep a read-only view of what every SaaS line item costs tend to catch the scraping bill drifting long before it shows up in a quarterly review. Data collection is one of the few line items that can quietly triple because a target site changed a setting.

Questions buyers ask

How much does ScraperAPI cost? Plans run from 49 USD a month for 100,000 credits to 1,975 USD for 21,500,000, with custom enterprise pricing above that and 10 percent off for annual billing. There is a 7 day trial with 5,000 credits and no card required. What you actually pay per page depends on the credit multiplier for your targets, not on the plan name.

What is a ScraperAPI credit? A credit is the unit of billing, and one normal page request spends one of them. Difficult categories cost more: 5 for large marketplaces, 25 for Google and Bing, 30 for LinkedIn. Parameters add on top, with JavaScript rendering and premium proxies each adding 10, and an anti-bot bypass adding 10.

How many credits does a Google search cost? 25 credits per results page, and the SERP category covers all Google and Bing subdomains. On the 49 USD plan that is 12.25 USD per 1,000 searches, which is why a daily rank tracking job is usually the line item that breaks a Hobby budget first.

Does ScraperAPI charge for failed requests? No. Their documentation says billing applies to successful requests, meaning status 200 and 404, plus requests cancelled from your side before the 70 second window closes. Ask the same question of any vendor you shortlist, because on hard targets a failure-billed plan and a success-billed plan are not comparable.

What is the cheaper alternative to ScraperAPI? It depends on your average credits per page, which is the number the comparison actually turns on. Below roughly 2 credits per page at entry level, staying on credits is cheaper. Above it, a flat per request API costs less, and the more your crawl leans on search results, marketplaces or protected pages, the wider that gap gets. Our web scraping API pricing is flat by design for exactly that reason, the Apify pricing breakdown covers usage based billing, Bright Data pricing per 1,000 pages covers the per gigabyte model, Oxylabs pricing per 1,000 pages covers the per target rate card, and the open source comparison covers the third option of running it yourself.

Is a scraping API worth it at all? If a developer is spending more than a few hours a month on proxies, headless browsers and block handling, the answer is usually yes at any of these prices. The build-versus-buy arithmetic, including what an engineer's time on maintenance really costs, is in the competitor price scraping software comparison.