ScraperAPI Review 2026: Pricing, Free Plan & Credit Costs
ScraperAPI is a web scraping API. You send it a URL, and it fetches the page through its rotating proxy pool, renders JavaScript if you ask, handles CAPTCHAs and bot protection, and returns the HTML or parsed JSON. You pay in API credits, and a single request costs anywhere from 1 to 75 of them depending on the site and the options you turn on.
That last point decides whether ScraperAPI is cheap or expensive for you, so this review spends most of its time on it: the seven plans, the free tier, the credit multipliers, and the workloads where a different tool costs less.
TL;DR
ScraperAPI is a good fit if: you already have scraper code (Requests, Scrapy, Playwright) that gets blocked, and you want one endpoint that handles proxies, retries and JavaScript rendering. It is cheapest on plain HTML pages, where a request costs 1 credit.
It is a poor fit if: most of your targets are JavaScript-heavy, behind Cloudflare or DataDome, or on Google and LinkedIn. Those requests cost 10 to 30 credits each, so a "1 million credit" plan buys far fewer pages than it sounds. It is also a poor fit if you need country targeting outside the US and EU on a small budget, or want ready-made scrapers and hosted runs without writing code.
Verdict: a solid, well-documented scraping API that has grown well beyond a proxy layer in 2026 (structured data endpoints, scheduled DataPipeline jobs, a crawler, an MCP server). Before you pick a plan, estimate your credit multiplier. It matters more than the price on the pricing page.
What ScraperAPI Actually Is
ScraperAPI's core product is a single HTTP endpoint. A basic request is a GET with your key and the target URL as query parameters:
curl "https://api.scraperapi.com?api_key=YOUR_KEY&render=true&country_code=us&url=https://example.com/page"
Or from Python:
import requests
params = {
"api_key": "YOUR_KEY",
"render": "true", # JavaScript rendering, 10 credits
"country_code": "us", # geotargeting, no extra credits
"url": "https://example.com/page",
}
html = requests.get("https://api.scraperapi.com", params=params, timeout=70).text
ScraperAPI's docs recommend a 70-second client timeout. Most responses arrive within seconds, but hard domains can take up to 60. They also ask you to put ScraperAPI parameters before url, so they don't collide with query strings already in the target URL. There is a proxy-port mode too, if you'd rather point an existing HTTP client at ScraperAPI as a proxy.
What's included on every paid plan, according to ScraperAPI's pricing page:
- Rotating proxy pools, plus premium residential and mobile IPs
- JavaScript rendering (headless browser)
- CAPTCHA handling and anti-bot bypass (Cloudflare, DataDome, PerimeterX, Akamai)
- Automatic retries, sticky sessions (
session_number, 15 minutes), desktop and mobile user agents - Unlimited bandwidth: you pay per request, not per GB
- Structured Data endpoints that return parsed JSON for Amazon, Walmart, eBay, Home Depot, Google (Search, News, Jobs, Shopping, Maps), Redfin, and AI answer engines (ChatGPT, Perplexity, Gemini)
- Async API for batch jobs with callbacks
- DataPipeline, a low-code job runner: up to 100,000 URLs, keywords, ASINs or Walmart IDs per project, input by form, CSV or webhook, output as HTML, JSON or CSV, with scheduling, webhook delivery and email status updates
- Crawler, which follows links from a seed URL and skips duplicates
- Integrations: an MCP server (including Claude Code), LangChain, LlamaIndex, n8n and Zapier
In 2026 ScraperAPI also acquired Traject Data, the company behind the Rainforest and SerpWow APIs, and moved those SERP and e-commerce data APIs onto the same credit system.
What you still don't get: a place to run your own code. Your scraper still runs on your servers. ScraperAPI also has no marketplace of community-built scrapers and no long-term storage of your results. Data goes to your webhook or your database.
Mental model: ScraperAPI is the fetch-and-unblock layer for your code, plus parsers for a short list of high-demand sites. You still own the scraper logic for everything else, along with the scheduling of your own code and the storage.
ScraperAPI Pricing 2026
Prices taken from ScraperAPI's pricing page in October 2026. Annual billing is 10% cheaper.
| Plan | Monthly | API credits | Per 1K credits | Concurrent threads | Geotargeting |
|---|---|---|---|---|---|
| Free | $0 | 1,000 | n/a | 5 | n/a |
| Hobby | $49 | 100,000 | $0.49 | 20 | US & EU only |
| Startup | $149 | 1,000,000 | $0.15 | 50 | US & EU only |
| Business | $299 | 3,000,000 | $0.10 | 100 | Global (country-level) |
| Scaling | $475 | 5,000,000 | $0.095 | 200 | Global |
| Professional | $975 | 10,500,000 | $0.093 | 300 | Global |
| Advanced | $1,975 | 21,500,000 | $0.092 | 500 | Global |
| Enterprise | Custom | 22M+ | Custom | 500+ | Global |
Every paid plan has the same features. What changes between tiers is credits, concurrency, geotargeting, how much analytics history you keep (30 days on Hobby and Startup, unlimited from Business), and support. Four billing rules are easy to miss:
- Credits don't roll over. The balance resets at every renewal.
- Pay-as-you-go overage exists only from Scaling up. On Hobby, Startup and Business, running out means upgrading or contacting support.
- Going over your concurrency limit returns HTTP 429. You aren't charged, but the request is rejected, not queued.
- You pay only for successful requests (HTTP 200 and 404), plus requests you cancel before the 70-second window ends.
There is a 7-day no-questions-asked refund on paid plans, and you can cancel any time from the dashboard.
Does ScraperAPI have a free plan?
Yes. New accounts start with a 7-day trial with 5,000 API credits, no credit card required. After that there is a permanent free plan with 1,000 API credits and 5 concurrent connections. That's enough to check whether ScraperAPI gets through on your target sites, but not to run anything in production. On the free plan the Crawler is also limited to depth 1 (the seed page and its direct links) and can't run on a schedule.
How ScraperAPI credits work
This table explains most "ScraperAPI is expensive" complaints. Costs come from ScraperAPI's credit cost documentation:
| Request type | Credits per request |
|---|---|
| Normal page | 1 |
| Amazon, Walmart, eBay | 5 |
| Google, Bing | 25 |
| 30 | |
| Sites behind Cloudflare, DataDome, PerimeterX or Akamai (bypass) | +10 |
render=true (JavaScript) | 10 |
premium=true (residential/mobile IPs) | 10 |
screenshot=true | 10 |
premium=true + render=true | 25 |
ultra_premium=true | 30 |
ultra_premium=true + render=true | 75 |
country_code, session_number, device_type, output_format, autoparse | no extra cost |
So a 1,000,000-credit Startup plan buys 1,000,000 plain pages, 100,000 rendered pages, 40,000 Google searches, or about 13,000 ultra_premium rendered pages. To control this, check a URL in the Domain Cost Estimator in the dashboard before you scale. Then set the max_cost parameter on every request: the API returns a 403 instead of silently charging more than your cap.
Pros
1. Integrates with code you already have Swap your request URL for the ScraperAPI endpoint, or use proxy-port mode, and existing Requests, Scrapy or Playwright code keeps working.
2. Unblocking without running your own browsers Proxy rotation, retries, CAPTCHA handling and named bypasses for Cloudflare, DataDome, PerimeterX and Akamai come in one endpoint. You don't maintain a headless browser fleet.
3. You only pay for successful requests Only HTTP 200 and 404 responses consume credits. With a per-GB proxy you pay for the bandwidth of every blocked attempt too.
4. Flat price per request, regardless of page size Bandwidth is unlimited, so a 3 MB product page costs the same as a 30 KB one. For heavy pages this can beat per-GB residential proxies outright.
5. Parsed JSON for the most-scraped sites For Amazon, Walmart, eBay, Google and Redfin, the Structured Data endpoints return parsed fields, so you skip writing and maintaining selectors.
6. Clear, honest billing tools
The cost estimator, max_cost, spending caps on pay-as-you-go, and a 7-day refund make it hard to be surprised by a bill if you use them.
Cons
1. Credit multipliers make the headline numbers misleading "1,000,000 credits" sounds like a million pages. With rendering it's 100,000. On Google it's 40,000. Budget by multiplier, not by credits.
2. Global geotargeting starts at $299/month Hobby and Startup are limited to US and EU targeting. If you need Brazil, India or Japan, the cheapest option is Business.
3. No overage on the three smaller plans If you run out of credits mid-month on Hobby, Startup or Business, you have to upgrade. Pay-as-you-go starts at $475/month (Scaling).
4. Credits expire monthly Spiky workloads, such as one large crawl per quarter, pay for idle months.
5. Your code, your servers DataPipeline schedules ScraperAPI's own templates, but your custom scraper still needs hosting, scheduling, monitoring and storage that you run yourself.
6. Structured endpoints cover a short list of sites If your targets aren't on that list, you write and maintain the parser yourself.
Common Complaints About ScraperAPI, Checked
AI assistants and review sites keep raising the same objections. Here is how each one holds up against ScraperAPI's current pricing and docs.
| Complaint | Holds up? | Why |
|---|---|---|
| "High pricing" | Partly | Plain HTML is cheap: $0.10–0.15 per 1,000 pages from Startup up. Rendered and protected pages cost 10–30× that, which is where the reputation comes from. |
| "Complex credit pricing" | Yes, but it's documented | The multipliers are public and the dashboard estimates cost per URL. It's still more to model than a flat per-page price. |
| "Geolocation limits" | Yes | Country-level targeting outside the US and EU requires Business ($299) or higher. |
| "Limited free tier" | Yes | 1,000 credits a month after a 5,000-credit trial. Enough to test, not to run anything. |
| "Inflexible plans" | Partly | Credits don't roll over, and overage exists only from Scaling up. You can change plans at any time, though. |
| "Slow response times" | Sometimes | Simple pages return in seconds. Protected domains can take up to 60 seconds, which is why the docs recommend a 70-second timeout. Use the Async API for volume. |
| "Limited features" | Outdated | This was fair for the 2023-era proxy API. The 2026 product adds structured endpoints, DataPipeline scheduling, a crawler, AI-answer scraping and an MCP server. |
| "Limited proxy control" | Yes, by design | You choose country, session and IP tier, but not specific IPs, ASNs or cities. If you need that control, buy raw proxies. |
When ScraperAPI Wins
You have working scraper code that gets blocked in production. Your Scrapy or Playwright scraper works locally and fails on a server. Put ScraperAPI in front of it and the blocking becomes ScraperAPI's problem.
Your targets are mostly plain HTML. At 1 credit per page, Startup works out to about $0.15 per 1,000 pages, and you never pay for bandwidth.
You scrape Amazon, Walmart or Google and want JSON, not HTML. The Structured Data endpoints remove the most fragile part of the job, the parser.
You already run your own scheduling and storage. If Airflow, cron or a queue worker already drives your jobs, ScraperAPI slots in as the fetch layer without adding another platform.
When ScraperAPI Doesn't Win
You're starting from scratch and don't want to write a scraper. Building and maintaining scrapers takes weeks. Apify has more than 83,000 ready-made Actors in its Store. You can run one in the browser, schedule it, and download the dataset without writing code.
You want your own code hosted, scheduled and stored. Running a scraper on a schedule, keeping its results and alerting Slack when something changes still means building infrastructure around ScraperAPI. Apify runs your code in its cloud, with scheduling, webhooks and dataset storage included.
You need countries outside the US and EU on a small budget. ScraperAPI's global targeting starts at $299/month. Apify's residential proxies, with country selection, are available on every plan including Free, billed at $8/GB from your usage credits.
Most of your pages are JavaScript-heavy or heavily protected. At 10 to 30 credits per page, 100,000 rendered pages need the 1M-credit Startup plan, and protected sites push you to Business. Compare that with a browser-based scraper billed per compute unit, or with a pre-built Actor priced per result.
You'd rather manage proxies yourself. If you have the engineering time for rotation, retries and headless browsers, raw proxies billed per GB or per IP can cost less than a scraping API at high volume, especially for lightweight pages. See our best residential proxies roundup and the proxy rotation guide for what that setup involves.
You need enterprise-grade proxy infrastructure and compliance. For city- and ASN-level targeting, KYC-backed networks and SLAs, Bright Data is the stronger choice. See Bright Data vs ScraperAPI.
ScraperAPI vs Alternatives
ScraperAPI vs Apify
| Dimension | ScraperAPI | Apify |
|---|---|---|
| Core product | Scraping API + structured endpoints | Actor platform + 83,000+ ready-made scrapers |
| Your code runs | Your servers | Apify cloud (serverless) |
| Ready-made scrapers | Structured endpoints for about 10 sites | 83,000+ Actors in the Store |
| Scheduling | DataPipeline and Crawler (ScraperAPI's templates) | Any Actor, including your own code |
| Result storage | Delivered to you (webhook, download) | Datasets included, export to JSON/CSV/Excel |
| Pricing unit | Credits per request (1–75) | Usage credits: compute units, proxy GB, or per-result Actor pricing |
| Free tier | 5,000-credit trial, then 1,000 credits | $5/month in usage credits |
| Entry paid plan | $49 (Hobby) | $19 (Starter, includes $19 of usage) |
| Best for | Fetch-and-unblock for code you own | End-to-end scraping jobs |
Choose ScraperAPI if you have working code and need a reliable fetch layer with predictable per-request billing.
Choose Apify if you're starting from scratch, want a pre-built scraper for your target, or need hosting, scheduling and storage in one place. Our full Apify vs ScraperAPI comparison goes deeper on architecture and cost.
ScraperAPI vs Bright Data
| Dimension | ScraperAPI | Bright Data |
|---|---|---|
| Primary product | Credit-based scraping API | Proxy networks + Web Unlocker + scraper APIs |
| Geolocation | US/EU below Business, country-level above | Country, city, ASN |
| Unblocking | Named bypasses, +10 credits | Web Unlocker |
| Pricing model | Per-request credits, bandwidth unlimited | Per GB (proxies) or per request (Unlocker) |
| Best for | Developer teams that want one endpoint | Enterprise scale, fine-grained targeting, compliance |
Choose ScraperAPI for simplicity and predictable billing at small and mid scale.
Choose Bright Data for granular targeting and enterprise procurement. The full breakdown is in Bright Data vs ScraperAPI.
Pricing Deep Dive: When Does It Get Expensive?
These scenarios use October 2026 list prices and the credit table above. Each one picks the cheapest plan that covers the credits.
| Scenario | Credits needed | Cheapest plan | Cost per 1,000 pages |
|---|---|---|---|
| 50K plain HTML pages/month | 50,000 | Hobby, $49 | $0.98 |
| 500K plain HTML pages/month | 500,000 | Startup, $149 | $0.30 |
| 200K pages, half need rendering | 100K + 1M = 1.1M | Business, $299 | $1.50 |
| 100K Google searches/month | 2,500,000 | Business, $299 | $2.99 |
| 1M rendered pages, global targeting | 10,000,000 | Professional, $975 | $0.98 |
Two patterns stand out:
- The cost per page depends on the multiplier more than on the plan. The cheapest row and the most expensive row run on plans that differ by $150, but the cost per page differs about tenfold.
- Watch the Startup-to-Business step. In the rendering scenario, 1.1M credits miss Startup's 1M by 10%, and with no overage below Scaling, you pay $299 instead of $149. If you are close to a plan limit, add
max_costand trim rendering to the pages that need it.
For comparison: if your target already has a well-maintained Apify Actor (for example, the Google Search Results Scraper), check its per-result price on the Actor page against the $2.99 per 1,000 searches above. For more platforms side by side, see our web scraping pricing comparison.
The Honest Take
ScraperAPI is a mature, well-documented scraping API. In 2026 it is a lot more than the proxy layer it started as: parsed endpoints for the most-scraped sites, scheduled low-code jobs, a crawler and an MCP server for AI agents.
The thing to understand before you buy is the credit multiplier. On plain HTML, ScraperAPI is one of the cheaper ways to fetch pages without getting blocked. On rendered, protected or Google pages, the same plan buys 10 to 25 times fewer pages, and the smaller plans have no overage to absorb a busy month.
If you have scraper code and a mostly-HTML workload, ScraperAPI is an easy recommendation. If you're starting from zero, want hosting and storage, or your target already has a ready-made scraper, a platform like Apify usually gets you to data faster and often for less.
Getting Started
- Start the trial: ScraperAPI free trial gives you 7 days and 5,000 credits, no card required.
- Price your targets: run each target domain through the Domain Cost Estimator in the dashboard to find its multiplier.
- Cap every request: add
max_costso a misconfigured request can't spend 75 credits. - Pick the plan by credits × multiplier: check concurrency and geotargeting at the same time.
- Re-check at month end: if you keep hitting the limit on Hobby, Startup or Business, compare the next tier with an Apify Actor for the same target.
Yes, if you have working scraper code and mostly plain-HTML targets. At 1 credit per page, the Startup plan works out to about $0.15 per 1,000 pages. It is less attractive when most pages need JavaScript rendering (10 credits), premium proxies (10), or are on Google (25) or LinkedIn (30), because the same plan then covers 10 to 30 times fewer pages.
Yes. New accounts get a 7-day trial with 5,000 API credits and no credit card required. After the trial, a permanent free plan includes 1,000 API credits with up to 5 concurrent connections. That is enough to test targets, not to run production scraping.
Paid plans are Hobby $49/month (100,000 credits), Startup $149 (1M), Business $299 (3M), Scaling $475 (5M), Professional $975 (10.5M) and Advanced $1,975 (21.5M), with custom Enterprise pricing above that. Annual billing is 10% cheaper. Credits reset monthly and do not roll over.
A normal page costs 1 credit. Amazon, Walmart and eBay cost 5, Google and Bing 25, and LinkedIn 30. JavaScript rendering, premium proxies or screenshots cost 10, premium plus rendering 25, ultra premium 30, and ultra premium plus rendering 75. Bypassing Cloudflare, DataDome, PerimeterX or Akamai adds 10. Only successful requests (HTTP 200 and 404) are charged.
Yes, with render=true, which costs 10 credits per request instead of 1. Use the Domain Cost Estimator to check each target, set max_cost to cap spend per request, and only enable rendering on pages that need it. For large rendered workloads, compare against a browser-based Apify Actor billed per compute unit or per result.
Yes, from the Business plan ($299/month) upward, at country level. Hobby and Startup are limited to US and EU targeting. Setting country_code costs no extra credits.
On Hobby, Startup and Business you upgrade to the next plan or contact support. On Scaling, Professional, Advanced and Enterprise you can keep scraping on pay-as-you-go at a fixed per-credit rate, with an optional monthly spending cap. Exceeding your concurrency limit returns HTTP 429 and is not charged.
Yes. ScraperAPI is an established scraping API with public documentation, published credit costs, a 7-day no-questions-asked refund, and billing only for successful requests. In 2026 it acquired Traject Data, the company behind the Rainforest and SerpWow APIs.
ScraperAPI is a scraping API: you send URLs from your own code and it fetches and unblocks them. Apify is a full scraping platform with 83,000+ ready-made Actors, hosted runs, scheduling and dataset storage. ScraperAPI suits teams with existing scraper code; Apify suits teams that want a ready-made scraper or don't want to run infrastructure. For SERP work specifically, the per-vendor cost comparison lives in best SERP scraping proxies.