Structured Data API

Try it

curl "https://api.rasterly.dev/v1/extract?url=https://shop.example/p/42&fields=title,price,currency,availability,brand,image,rating" \
  -H "X-Api-Key: sk_live_..."

{
  "url": "https://shop.example/p/42",
  "fields": {
    "title": "Acme Wool Runner",
    "price": "98.00",
    "currency": "USD",
    "availability": "InStock",
    "brand": "Acme",
    "image": "https://…/p1.jpg",
    "rating": "4.7"
  }
}

Fields with built-in synonyms

• Commerce — price, currency, availability, brand, sku, gtin, rating, ratingcount.
• Content — title, description, author, published, category, image.
• Page — url/canonical, sitename, lang.
• Anything else — any field name falls back to a case-insensitive lookup across the page's JSON-LD, OpenGraph and meta.

Absent fields come back null — deterministic honesty beats a confident guess.

What it's for

• Price & availability monitoring — poll competitor PDPs for price,currency,availability without writing a parser per site.
• Product feeds — normalize catalog pages into a flat record for a database or a shopping index.
• Agent tool calls — an LLM agent gets the two fields it needs as JSON, instead of a page of Markdown to re-parse.
• Content ingestion — pull title,author,published,image for a reading list or link preview.

Why rasterly

It's the same one-render engine as the screenshot and read endpoints, so it's priced from $0.90 per 1,000 calls with 100/mo free. Because field mapping is deterministic — no model in the loop — you skip the per-call token cost that generic LLM-extract APIs pass on, and you never get a hallucinated price. Need the full payload instead? Drop fields= and /v1/extract returns all JSON-LD, tables, images and links.

FAQ