The DataSonar API

The Universal Scrape Endpoint

This is the workhorse of the platform. Send us a URL, and we'll return the fully rendered HTML or Markdown. Need to click a button, type into a search box, or accept a cookie banner before extracting? Use our Native Actions Pipeline.

The Native Actions Pipeline

Instead of writing custom JavaScript to manipulate the DOM, you can pass an array of human-readable actions. These are compiled directly into our underlying browser engine, executing natively without the overhead of the CDP protocol.

{
  "url": "https://example.com/login",
  "format": "html",
  "stealth": true,
  "actions": [
    { "action": "type", "selector": "input[name='email']", "text": "agent@company.com" },
    { "action": "type", "selector": "input[name='password']", "text": "secure123" },
    { "action": "click", "selector": "button[type='submit']" },
    { "action": "scroll_to_bottom" }
  ],
  "js_eval": "document.title"
}
          

Scale seamlessly with Batch & Async

• POST /v1/scrape/batch

  Pass an array of up to 100 URLs. We will scrape them concurrently using our worker pool and return a unified result set.

• POST /v1/scrape/async

  For massive jobs, queue the work. We'll return a Job ID immediately and ping your webhook when the extraction finishes.

• POST /v1/scrape/smart

  Not sure if a page needs full browser rendering? Smart scrape evaluates the target. If it's a static site, we bypass the browser and fetch it instantly over raw HTTP, saving you time and bandwidth.