Extract Data

Uses AI to extract specific structured data from a web page. Define the fields you want, and the AI finds and extracts them.

Type: TINY_EXTRACT Color: Orange (#F97316) Credits: 2 per run Tabs: Initialise → Configure → Test

Templates

Template	Schema fields	Use case
Extract Pricing	plan_name, price, features	Competitor pricing pages
Extract Contacts	name, email, phone, role	Contact/team pages

Configure tab fields

Field	Type	Required	Description
URL	FX formula	Yes	The page to extract from
Schema	Array	Yes	Field definitions: what to extract
Prompt	String	No	Natural language instruction for the AI

Schema fields

Each field has:

Key: field name (e.g., price, company_name)
Description: what to look for (e.g., "Monthly price in USD")

At least one key must be non-empty.

Output

JSON object matching your schema:

{
  "plan_name": "Pro",
  "price": "$49/month",
  "features": ["Unlimited users", "API access", "Priority support"]
}

Example: Extract product details

Schema:

product_name: "The main product name"
price: "Current price including currency"
rating: "Average customer rating"
review_count: "Number of reviews"

The AI reads the page and fills each field.

Tip

Be specific in field descriptions. Instead of "price", write "Monthly subscription price in USD, excluding taxes." The more specific, the more accurate the extraction.