Starter
49 USD 24 USD
a month, 288 USD billed once a year, save 50%a month, billed monthly
- Requests per month
- 50,000
- Concurrent requests
- 5
Scraping API
Web Scraping API with Rotating Proxies and JS Rendering
Send one request. Get the page, unblocked, and the fields you asked for.
Billed only on successful requests. Your data stays yours.
- Status
- ...
- Elapsed
- ... ms
- Exit country
- ...
- Browser used
- ...
This console sends a real request and shows you the real response. Five live fetches per session, and the sample targets run as often as you like.
WebScrap fetches any public URL through rotating proxies, renders it in a real browser when the site needs one, handles the block, and returns the fields you named as typed JSON. You pick a plan by monthly requests and stop maintaining a scraper.
The answer is the response body, not a quote
Here is the request you just composed and the payload that came back. An engineer believes a response, so that is what this page shows instead of opinions about the product.
The request
Copy this into your own repository and it works the same way from your machine as it does from this page.
The response
Status, elapsed milliseconds, exit country and the typed JSON built from the page you targeted. No page left for you to parse afterwards.
Stop maintaining a scraper
The proxy pool, the browser farm, the retry logic and the parser live on our side and stay our problem when a site ships a redesign.
Stop getting blocked
Fingerprints, headers, sessions and retries rotate on every request, so a page that refuses a plain fetch comes back with a 200 through the web scraping proxy layer.
Know the bill before the month starts
A plan is a monthly request limit at a fixed price, billed only on successful responses, so finance gets one predictable line item instead of an open meter.
One request, four steps
-
1
Request: you call one endpoint with a URL, a country, a render mode and the fields you want back.
-
2
Proxy and render: we pick an exit address in the country you asked for and load the page in a real browser when the site needs one.
-
3
Unblock and retry: challenges, redirects and rate limits are handled in the request, with a new identity on every retry.
-
4
Parse: your CSS, XPath or plain English schema turns the page into typed JSON before it leaves our side.
Every step is described in detail on how web data scraping works.
Read the docsTry it on a page you care about
Paste any public URL, or start from one of the four targets below. Switch Render to Browser on the JavaScript heavy target and watch the payload change, which is the whole difference between a plain fetch and a rendered one.
Schema mode takes CSS selectors and runs in your browser at no cost. Describe it takes plain English, for example product name, price, currency, availability, and returns typed JSON. Raw HTML returns the document exactly as the proxy saw it.
- Status
- ...
- Elapsed
- ... ms
- Exit country
- ...
- Browser used
- ...
Three ways to get the data, and what each one costs you
| Question | Build it yourself | Run a headless cluster | WebScrap |
|---|---|---|---|
| Engineering time to first data | Weeks, proxies and parsers first | Days, plus cluster setup | Minutes, one HTTP call |
| Time lost per site redesign | Every parser is rewritten by hand | Every parser is rewritten by hand | Change the schema in the request |
| Cost when a site changes its anti-bot | An unplanned sprint | An unplanned sprint plus more nodes | Included, handled on our side |
| Cost model | Servers, addresses and salaries | Servers, addresses and salaries | A monthly request limit at a fixed price |
| Who is on call at 3 a.m. | Your engineer | Your engineer | We are |
The honest version of this table is on the open source scraping tools comparison, including the cases where building it yourself is the right call.
What a VP Engineering or CTO gets on paper
The question upstairs is never which parser library is nicer. It is what this costs next quarter, who is accountable when it breaks, and what happens when the engineer who wrote the scraper leaves.
- A predictable monthly line item instead of an infrastructure project with no end date
- Single sign on through SAML or OIDC, SCIM provisioning and custom roles on Enterprise
- An audit log of every API key, every schema and every scheduled job change
- A 99.9 percent uptime commitment and a four business hour response on Scale
- An invoice for every period, a security questionnaire answered in writing, a DPA and data residency in the EU or the US
- Responses are never resold, and the data you collect stays yours
Everything a security review asks for is set out on the enterprise web scraping page.
Compare plansPlans by monthly requests, nothing else to count
Every plan is a monthly subscription with a request limit, billed only on successful responses. Yearly billing costs exactly half of twelve monthly payments, you save 50% on every plan. There is no metered overage bill waiting at the end of the month.
Growth
Recommended
149 USD 74 USD
a month, 888 USD billed once a year, save 50%a month, billed monthly
- Requests per month
- 250,000
- Concurrent requests
- 50
Scale
499 USD 249 USD
a month, 2,988 USD billed once a year, save 50%a month, billed monthly
- Requests per month
- 1,500,000
- Concurrent requests
- 200
Enterprise
998 USD 499 USD
a month, 5,988 USD billed once a year, save 50%a month, billed monthly
- Requests per month
- 3,000,000
- Concurrent requests
- 400
Estimate your monthly cost
Enter how many requests you plan to send in a month and the share you expect to come back successful. Only successful responses count against the plan, so that number picks the plan for you.
%
- Successful requests per month
- 108,000
- Recommended plan
- Growth
- Monthly price
- 149 USD998 USD
Cost per 1,000 successful requests
1.38 USD
Enterprise, the largest plan, includes 3,000,000 successful requests a month for 998 USD. Requests above the limit return a quota error, never a surprise invoice.
The estimate uses the list prices on this page and nothing else.
The full grid with every feature per plan is on the best web scraping api pricing page.
Is web scraping legal
Reading public pages is generally lawful in the jurisdictions our customers operate in, and courts have repeatedly declined to treat access to publicly available data as unauthorised access. What matters is what you collect and what you do with it: personal data carries obligations under GDPR and similar laws, and a contract you accepted can bind you further. We wrote the longer answer in is web scraping legal, and it is not legal advice.
Will my requests get blocked
Some will, on the first attempt. The API retries with a new address, a new fingerprint and a browser when the page needs one, and you are billed only for the response that actually comes back. Blocks handled inside the request never reach your code.
What happens when a site changes its layout
Your schema breaks, not your integration. You change the selector or the plain English description in the request body and the next call returns the corrected fields. There is no parser to redeploy on your side.
Do I pay for failed requests
No. The counter moves on successful responses only. Timeouts, blocks and retries are on us, which is why the monthly number is knowable before the month starts.
Who owns the data I collect
You do. We hold responses only as long as the retention window of your plan, we never resell them, and we never use your targets or payloads to build anything for anyone else.
Can I use it from Python or without writing code
Both. There is a python scraping tool quickstart with requests and pandas, and a no code web scraper with a visual schema builder and scheduled jobs that post to a spreadsheet or a webhook.
What happens to my data after the request
Responses are deleted when the retention window of your plan ends: 24 hours on Starter, 7 days on Growth, 30 days on Scale. Deletion covers the payload, the screenshot and the rendered HTML.
Create your account and run your first request
One field, one minute. Enter your work email, confirm the six digit code and your account is registered.