Built-in Templates

Templates are ready-made extractors for the things people scrape most often. Add a templates array to any request and the results come back under data.template_extracted — no CSS selectors required.

Available templates

Template	Extracts
links	All anchor URLs on the page
images	All image URLs
headings	Heading text (h1–h6), grouped by level
emails	Email addresses found in the page
phone_numbers	Phone numbers found in the page
metadata	<meta> tags, title, Open Graph data
tables	HTML tables parsed into rows and columns

Example

curl -X POST https://api.omniscrape.io/v1/scrape \
  -H "X-API-Key: $OMNISCRAPE_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://news.example.com",
    "templates": ["links", "images", "headings", "emails"]
  }'

{
  "data": {
    "template_extracted": {
      "links": ["https://news.example.com/a", "https://news.example.com/b"],
      "images": ["https://cdn.example.com/1.jpg"],
      "headings": { "h1": ["Top Stories"], "h2": ["World", "Tech"] },
      "emails": ["press@example.com"]
    }
  }
}

Combining templates with other features

Templates work alongside everything else — modes, proxies, and CSS extraction. For example, get a contact email and the company name in one call:

{
  "url": "https://company.example.com/about",
  "templates": ["emails", "phone_numbers"],
  "css_selectors": { "company": "h1.company-name" }
}

You'll receive both data.template_extracted and data.css_extracted.

Templates vs CSS extraction

• Use templates for generic, page-wide collections (every link, every image, all emails).
• Use CSS extraction when you need specific, named fields from known elements (the price, the title, the rating).

Tables

The tables template is handy for scraping data grids and spec sheets — each table becomes a structured array of rows, so you can drop it straight into a spreadsheet or database without parsing HTML.