Billing & credits

ScrapeIt bills in credits. Every Web Scraping API response reports what the request cost:

{
  "cost": { "estimated_credits": 4, "charged_credits": 4 }
}

• cost.estimated_credits - the price computed before the request ran.
• cost.charged_credits - the credits charged for the request.

A site API response reports what the request charged next to its data, in meta:

{
  "data": { "...": "..." },
  "meta": { "cost": { "charged_credits": 30 } }
}

What a request costs

What the request uses	Credits
Fetch + datacenter proxy	4
Browser render + datacenter proxy	30
Fetch + residential proxy	15
Browser render + residential proxy	50
proxyType: "auto"	the price of the proxy tier that served the request
Screenshot	the render price, nothing on top
AI extraction	+40
Site APIs (Amazon, Walmart, Google SERP, Google Maps, Zillow, Allegro, Booking.com, Yelp, Redfin, Airbnb, Tripadvisor, Yellow Pages, eBay, Glassdoor, Apartments.com)	one flat price per request, 15 to 50 - proxy and render included
Site API with include_html: true (the page HTML in html)	+4
Cache hit	half of the price, rounded up

Prices are flat per request. The prices of the newest site APIs (Google Maps, Booking.com, Amazon Best Sellers, Redfin, Airbnb, Tripadvisor, Yellow Pages, eBay, Glassdoor) are introductory and may change.

Details and examples: Request cost. The Playground cost preview and cost.estimated_credits in every response show the estimate next to the final charged_credits.

Plans

Every plan grants a monthly credit allowance and a concurrency limit. Annual billing is 20% cheaper than paying monthly. The allowance renews on the same day each month as your period started; in a month without that day (the 29th to 31st), on its last day.

Renewal tops your balance up to the monthly allowance, it does not add the allowance on top: unused credits carry over only up to the allowance. A balance already at or above it stays as it is and gets nothing that month. An upgrade in the middle of a month adds the difference between the two allowances.

Prices exclude VAT. Checkout adds it by your billing address and tax ID: customers in Poland and consumers in the EU pay VAT, an EU business with a valid VAT ID is reverse-charged, and customers outside the EU pay none.

Plan	Price per month	Credits per month	Concurrent requests	URLs per batch	Raw result retention
Free	€0	3 000	2	100	24 hours
Starter	€35	360 000	5	5 000	3 days
Pro	€79	2 000 000	16	10 000	7 days
Business	€169	5 000 000	25	25 000	14 days
Ultra	€449	15 000 000	40	50 000	21 days

The free credits go to the first workspace of an account only. A workspace you create later starts on Free with an empty balance and gets no monthly free credits; a paid plan grants its credits to it as usual.

Every plan includes the Web Scraping API, batch scraping, the Website Crawler and webhooks; AI extraction is available on paid plans. Your plan's current values are always shown on the dashboard Pricing and Usage pages.

Shared cache discount

A request served from the shared cache costs half the normal credits.

When a request is charged

A request is charged once the target website answered and the result was saved - whatever the website answered:

• its data;
• its own HTTP 4xx/5xx error page (an "access denied" page included) or an empty page - that is the website's real answer.

A request is not charged when it fails on our side before a result: proxy, network, browser, internal and timeout failures, and an anti-bot challenge every retry still met (anti_bot_challenge) - no content reached you. Retries through other proxies inside one request are never charged one by one: you pay for the request, not for its attempts. Requests refused before they run cost nothing either: VALIDATION_ERROR, an invalid API key, insufficient credits, CONCURRENCY_LIMIT_EXCEEDED, URLs blocked by the private-address (SSRF) policy and requests to a domain paused for answers over the response size limit (response_too_large).

Site APIs, no-code and bulk runs go further and give credits back when the target blocked them - see the next section.

When credits come back

A site API request, a no-code run or a bulk run that ends without data through no fault of yours gives back every credit it charged:

• the target answered only with bot checks (502 TARGET_BLOCKED);
• your workspace's requests to the site were paused because its answers kept exceeding the size limit (502 TARGET_RESPONSE_TOO_LARGE);
• a later step failed on our side: no browser capacity, a timeout, no proxy, a restart of the service, our DNS resolver not answering (503 TARGET_DNS_UNAVAILABLE) or an internal error (500 INTERNAL_ERROR, 503 SERVICE_UNAVAILABLE);
• a background no-code or bulk run was cut off by a restart of the service: it ends failed, and its credits come back.

A page the target answered stays charged, even when it holds no data for you: a not-found page (404 NOT_FOUND), the detail page of a card the target blocked inside a listing that succeeded (with include_detail that card comes back without detail), and the page of a bulk run blocked after earlier pages returned data.

Monitoring your spend

• Dashboard Usage page - per-day and per-product breakdown.
• GET /api/v1/usage/me - current period usage for dashboards of your own.
• Batch and crawler jobs expose credits_spent per job and per item, so you can attribute cost to a single URL.

Related pages

• Concurrency & limits
• Web Scraping API - which options affect price.
• FAQ - the questions everyone asks first.