scrapebento Public API - Scrapebento Docs

scrapebento is a public API for agent-facing search, scraping, and brand extraction workflows. It is designed for automations that need structured data, optional synchronous responses, and a consistent async job model when work takes longer.

What you can do

Search Google and return only the SERP payload with /v1/public/search.
Ask grounded questions over the top result documents with /v1/public/search/answer.
Scrape a single URL as markdown, raw HTML, JSON, metadata, or screenshot with /v1/public/scrape.
Extract raw site signals with /v1/public/scrape/site, a brand visual-identity profile (palette, typography, logo, footer) with /v1/public/scrape/brand-design, or a heavier company profile with /v1/public/scrape/company.
Run domain and community intelligence workflows across mentions, SEO, infra, ads, social search, social engagement, and feeds.

How responses work

Every successful response uses the repo’s data envelope:

{
  "data": {
    "status": "succeeded",
    "kind": "url",
    "creditsCharged": 1,
    "result": {}
  }
}

If work does not finish within the sync wait window, the API returns 202 Accepted with a jobId and pollUrl. Failures use the shared error envelope for immediate request validation/auth problems, or a terminal job payload with status: "failed" for scrape/search execution failures.

Start here

Quickstart for the first bearer-token calls.
MCP server for connecting an agent over stateless Streamable HTTP.
Auth & Credits for scopes, metering, cache, and 402 behavior.
Jobs & Webhooks for polling, sync/async behavior, and callback delivery.
Grounded Search for /search vs /search/answer.

​What you can do

​How responses work

​Start here

What you can do

How responses work

Start here