WebRobot

ScrapingAnt Pricing for Every Plan and What a Request Costs in Credits

6 min read

ScrapingAnt pricing runs from $19 a month for 100,000 API credits to $599 for 8,000,000, with a Custom tier from $699. The catch is the meter: a page costs 1 credit on a datacenter proxy and 125 on a residential proxy with JavaScript rendering, so the $19 plan buys anywhere from 800 to 100,000 pages. This page prices both ends and three real jobs.

Run the robot

figures taken from the ScrapingAnt pricing section and its credit cost documentation; prices in USD before tax

Robot console

No account needed Running · s Complete · rows

Or try:

Reads one page, up to 25 rows, usually in under a minute.

Agent log

▌

Extracted data ·

The page was read, but none of the data you described is on it. Try describing it the way it reads on the page, or another page.

more rows found on this page. Confirm your email to see every row and download the CSV.

Want this data fresh every morning? A plan runs this robot on a schedule, follows Next pages and feeds Sheets and the API.

The plans

ScrapingAnt plans, prices and credits at a glance

ScrapingAnt publishes five tiers on its pricing section. The columns on the right turn credits into pages for the two requests that matter most: a rendered page on the standard proxy (10 credits) and a rendered page on the residential proxy (125 credits).
Plan Monthly Credits a month Pages at 10 credits Pages at 125 credits Per 1,000 pages at 10 Per 1,000 pages at 125
Free $0 10,000 1,000 80 $0 $0
Enthusiast $19 100,000 10,000 800 $1.90 $23.75
Startup $49 500,000 50,000 4,000 $0.98 $12.25
Business $249 3,000,000 300,000 24,000 $0.83 $10.38
Business Pro $599 8,000,000 800,000 64,000 $0.75 $9.36
Custom from $699 10,000,000 or more 1,000,000 or more 80,000 or more quote quote

Enthusiast support is by email and integration help is docs only. Startup adds priority email and custom code snippets, Business adds debug sessions, and custom proxy pools and anti-bot work start at Business Pro and Custom. Unused credits do not carry over. If you run dry mid month, ScrapingAnt lets you restart the plan and refreshes the credits at once, which is convenient but means a second charge.

The meter

How many credits a ScrapingAnt request costs

This is the part that moves the bill, and it comes from the credit cost documentation. Failed requests cost nothing.
Request settingsCreditsPages on the $19 plan
Simple request, no browser, standard proxy1100,000
Headless browser without JS rendering, standard proxy250,000
Headless browser with JS rendering, standard proxy1010,000
Simple request, residential proxy254,000
Headless browser without JS rendering, residential proxy502,000
Headless browser with JS rendering, residential proxy125800
Any request to a Google domain, standard proxy1010,000

The AI extractor is billed on top. It charges 1 credit for every 30 characters of the page as Markdown plus the output text, and then adds the request cost. A product page that comes to 20,000 characters of Markdown with 1,000 characters of output costs 700 credits for the extraction plus 10 for the render, so 710 credits. On the $19 plan that is about 140 pages, not 10,000. The headline credit count says little until you know which mode your target site forces.

Worked bills

Three real jobs priced on ScrapingAnt and WebRobot

The ScrapingAnt figures use the credit table above. The WebRobot figures are the published WebRobot pricing: Launch at $79 a month for 10,000 records and Scale at $249 for 100,000.
JobCredits neededScrapingAnt planWebRobot planWhat ScrapingAnt leaves to you
10,000 detail pages a month on a site that renders with JavaScript but accepts datacenter IPs 100,000 Enthusiast, $19 Launch, $79 Parser, schedule, retries, storage
10,000 detail pages a month on a protected retailer that needs residential plus rendering 1,250,000 Business, $249 (3,000,000 credits) Launch, $79 The same, plus tuning requests that fail
3,000 protected pages every day, 90,000 a month 11,250,000 Custom, $699 or more Scale, $249 (100,000 records) The same, at ten times the volume

Read the first row honestly: if you have a developer, a rendered but unprotected site and 10,000 pages, ScrapingAnt at $19 is the cheaper way to fetch HTML and WebRobot does not win on price. The gap closes as soon as the target blocks datacenter IPs, because the residential multiplier is 12.5 times, and it reverses when the person who needs the data cannot write a parser. WebRobot counts finished records, not requests, so a retry or a heavier proxy does not change the invoice.

Which plan

Which ScrapingAnt plan to buy, or whether to skip it

  • Enthusiast ($19) suits one developer fetching up to 10,000 rendered pages from sites that do not block datacenter ranges.
  • Startup ($49) has the best price per page of the small plans and covers 50,000 rendered pages or 4,000 residential ones.
  • Business ($249) pays off when you pass roughly 500,000 credits a month and want debug sessions.
  • Skip it if nobody on the team writes code, or if the site you need sits behind heavy bot protection and a residential plan would eat the pool.

The no-code route

WebRobot runs its own browsers and proxies, so there is no credit meter and no parser. You paste a URL and write the columns you want, such as "product, price, stock, rating". The robot follows pagination and detail pages, runs daily on Launch or hourly on Scale, and delivers the rows to a sheet. If you are weighing several credit APIs, the ScrapingBee pricing and Scrape.do pricing breakdowns use the same worked-bill format, and the ScrapingBee alternative page shows the switch step by step, and the ScrapingAnt alternative page does the same for this vendor. The web scraping API page covers the developer route when you do want JSON over an endpoint.

FAQ

ScrapingAnt pricing questions buyers ask

ScrapingAnt sells five plans: Enthusiast at $19 a month for 100,000 API credits, Startup at $49 for 500,000, Business at $249 for 3,000,000, Business Pro at $599 for 8,000,000 and a Custom tier from $699 for 10,000,000 or more. Every account also gets 10,000 free credits a month with no card. What a page costs depends on the request settings, from 1 to 125 credits.

A simple request on a datacenter proxy costs 1 credit and a headless browser with JavaScript rendering costs 10. On the premium residential proxy the same requests cost 25 and 125 credits. Any request to a Google domain costs 10 credits on the standard proxy. The AI extractor adds 1 credit for every 30 characters of page Markdown and output.

Yes. The free plan gives 10,000 API credits every month with no credit card required. At the default browser request of 10 credits that is 1,000 rendered pages, or 10,000 plain datacenter requests. Unused credits do not roll over to the next period, so the free pool is for testing a target site, not for a standing job.

No. The ScrapingAnt credit documentation states that unused API credits are not transferred to the next subscription period. If you run out mid-month you can restart the plan and the credits refresh the moment you pay, but a half-used pool is gone at renewal.

No. ScrapingAnt says you pay only for successful scrapes and that failed requests cost 0 credits. Each response carries an Ant-credits-cost header with the credits that request used, so you can check your own bill against the table in the credits documentation.

On raw page fetches, often yes. A plain 10,000 page job fits the $19 plan. But ScrapingAnt returns HTML, and someone still writes the parser, the schedule and the storage. When a non-developer needs finished rows in Sheets every day, a flat plan such as WebRobot Launch at $79 for 10,000 records removes that work.

Skip the parser and get the rows

Describe the columns, check a live run, and get data in Sheets, Excel or the API every day. Plans from $79 a month.