Scrapinator
Documentation

Live in 60 seconds.

Get an API key. Hit one endpoint. Get clean structured data back. The full reference is being polished — here's the quickstart.

Try 500 credits freeSee every surface
1

Authenticate

Every request needs your API key as a Bearer token.

envbash
export SCRAPINATOR_KEY="sk_live_••••"
2

Make your first request

One endpoint per surface. Same body shape across all of them.

quickstart.tsts
import { Scrapinator } from "scrapinator";

const sc = new Scrapinator({ apiKey: process.env.SCRAPINATOR_KEY });

const { result } = await sc.scrape({
  surface: "chatgpt",
  prompt: "best CRM for early-stage startups",
  country: "US",
});
3

Handle the response

Every surface returns the same shape: markdown, sources, queries, entities, and the raw capture URLs.

result.jsonjson
{
  "markdown": "### Best CRMs ...",
  "sources": [
    {
      "url": "https://...",
      "title": "...",
      "cited_text": "..."
    }
  ],
  "queries": [
    "best crm 2025"
  ],
  "entities": [
    "Attio",
    "HubSpot"
  ]
}
4

Re-parse a saved capture — free

When your schema changes, hit re-parse on any saved capture. It costs zero credits.

reparse.tsts
await sc.reparse({ captureId: "cap_2W5x...", schemaVersion: "v3" });

Need the full reference?

Drop us a line — we'll send you a sandbox key and the in-progress reference.

Try 500 credits freeTalk to us