Overview
The Interact surface drives a real Chromium browser. You provide a starting URL and a sequence of actions; the API executes them in order and returns the page state after the last action.Basic example: fill and submit a form
Action types
| Type | Required fields | Description |
|---|---|---|
goto | url (optional) | Navigate to a URL. Omit url to reload the current page. |
click | selector | Click an element matching the CSS selector. |
type | selector, value | Focus the element and type the value. |
select | selector, value | Set a <select> element’s value. |
press | key | Press a keyboard key (e.g. "Enter", "Tab"). |
scroll | selector?, direction?, amount? | Scroll the page or a specific element. |
wait_for | selector | Wait until an element matching the selector appears in the DOM. |
wait_ms | duration_ms | Pause for a fixed number of milliseconds. |
Taking a screenshot
Include"screenshot" in output to capture the page after the last action:
outputs.screenshot.url — a signed URL valid for 24 hours.
Extracting data after interaction
Combine actions withextract to pull structured data after the browser is in the right state:
Persistent sessions
Reuse browser state across multiple Interact calls using sessions:Async interactions
Long interaction sequences should be run as async jobs:POST /v1/jobs and poll GET /v1/jobs/{id}. See Async Jobs.