Scraper APIPOST /v1/scrape

Send a URL. Get the page.

The Scraper API fetches any public page through residential proxies, renders JavaScript when the page needs it and retries around blocks. You get HTML or Markdown back and pay only when it works.

What it does

Any public page as HTML or Markdown, with proxies, rendering and retries handled.

  • HTTP first, browser when needed

    With render=auto a request starts on a fast HTTP client with a real browser fingerprint and moves to headless Chrome only if the page needs it. A second browser engine steps in when Chrome is blocked.

  • Residential proxies by country

    Pick the exit country with country, or keep the same IP for 10 minutes with session.

  • Blocks are retried, not returned

    Challenge pages from the common anti-bot vendors are detected and retried with a new IP, up to 3 attempts within your timeout.

  • HTML, raw or Markdown

    format=markdown returns the main content as clean Markdown, without navigation, footers or scripts. Good input for an LLM or a RAG index.

  • wait_for, screenshots and iframes

    Wait for a CSS selector before the page is captured, get a PNG of the viewport next to the HTML, and include the content of embedded iframes.

  • Async, batches and webhooks

    Send async: true and get a signed webhook when the job is done, or submit up to 1,000 URLs at once with Batch.

Example

One HTTP call

POST a URL with your key in the X-API-Key header. The answer has the target's status, headers and body, plus the tier, attempts and credits used.

Official SDKs are coming soon. Until then, any HTTP client works.

Scraper API docs
curl https://api.siftzo.com/v1/scrape \
  -H "X-API-Key: $SIFTZO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com", "country": "de"}'

Reference

Key parameters

urlstringrequired
A public http or https URL.
renderauto | never | alwaysdefault auto
never is HTTP only, always uses a browser, auto decides and escalates. See Rendering & tiers.
countryISO 3166-1 alpha-2
Exit country, for example de.
sessionstring
Keeps the same exit IP for 10 minutes.
formatjson | raw | markdown | raw_markdowndefault json
JSON with metadata, the body as-is, or the page as Markdown.
wait_forCSS selector
Browser only. Wait for this element.
screenshotbooleandefault false
Browser only. PNG.
iframesbooleandefault false
Browser only. Include the content of the page's iframes in the HTML.
asyncbooleandefault false
Return a job ID at once. Add webhook_url for a signed callback.
timeout_msintegerdefault 60000
5,000–90,000, for the whole request including retries.

Every parameter and response field is in the Scraper API docs.

Credits

What it costs

Scraper API credits
RequestCredits
HTTP request (render=never)1
Browser request (render=always, or auto when needed)10
Markdown output (format=markdown)no extra credits
Batch itemsame as the scrape it runs
Failed request (blocked, timeout, proxy error)0

You pay for the tier that produced the result, once per request. Only answers with 2xx, 3xx, 404 or 410 are billed.

See plans

Questions

When does a request use a browser?

With render=always, with wait_for, screenshot or iframes, or with render=auto when the HTTP attempt is blocked. Siftzo also remembers domains that need a browser and starts there next time, while checking now and then whether HTTP works again.

What does the Markdown output contain?

The main content of the page: headings, links, lists, tables, code blocks and images, with links made absolute. Navigation, footers, forms and scripts are removed. Markdown costs the same as JSON. See Markdown.

Do I pay for blocked requests?

No. Blocks, timeouts, unreachable targets and proxy errors cost 0 credits, including the retries Siftzo makes for you.

How long are async results kept?

24 hours. Fetch them with GET /v1/jobs/{id}, or have them sent to your webhook.

Try the Scraper API for free

Create an account, copy your key and send a request. No card needed for the Free plan.