Overview | OpenSERP Docs

OpenSERP Cloud is the hosted OpenSERP API for live search results from Google, Bing, Yandex, Baidu, DuckDuckGo, and Ecosia — through a single JSON envelope.

It is the same search engine as the open-source server, with a /v1/ prefix, API-key authentication, account credits, and hosted reliability layered on top. Already on OSS? Most code only needs a new base URL and an Authorization header — see OSS vs Cloud.

Try it before integrating

The Search Playground is the fastest way to see real results: pick an engine, run a query, inspect the response, and copy a working curl. No code required.

When you are ready to wire it into an app, the Quickstart takes about three minutes.

A two-line taste

import { OpenSERP } from "@openserp/sdk";

const client = new OpenSERP({ apiKey: process.env.OPENSERP_KEY });
const { results } = await client.search({ engine: "google", text: "openserp", limit: 5 });

That is the full happy path. The SDK adds the base URL, the Authorization header, query encoding, and typed responses for you.

Who Cloud is for

You are building	Cloud helps with
AI agents, RAG, or LLM tool-use	One API for fresh web results with a predictable JSON envelope
SEO monitoring, rank tracking, competitor research	Multi-engine coverage, pagination, region/language filters, usage headers
Search-powered product features	Hosted keys, credit billing, low setup time
Prototypes and internal tools	Pay-as-you-go credits, no monthly minimum

If you need full infrastructure control or zero per-call API cost, the open-source server is free under MIT and remains a first-class option.

What you can do today

Single-engine web search — /v1/google/search, /v1/bing/search, and friends.
Multi-engine merged search — /v1/mega/search?mode=balanced returns one deduplicated list.
First-success routing — mode=any or mode=fast returns one successful response for a predictable 1 credit.
Image search — single-engine and multi-engine variants.
Filters — language, region, date range, site restriction, file type, pagination.
Account & capability checks — /v1/me, /v1/pricing, /v1/engines/status, /v1/engines/capabilities.
Usage telemetry — X-Credits-Used, X-Credits-Remaining, X-Engine-Used headers on every billable response.

Official SDKs

Both SDKs target Cloud and the self-hosted server with the same code — only the constructor changes.

Runtime	Package	Install
TypeScript / JavaScript	`@openserp/sdk`	`npm install @openserp/sdk`
Python	`openserp`	`pip install openserp`

A Go SDK is on the roadmap. Until then, any HTTP client works.

Tools & integrations

Use OpenSERP from AI agents and no-code workflows without writing client code:

Integration	Package	Use it for
MCP server	`@openserp/mcp`	Give Claude, Cursor, and other MCP clients live search and URL extraction as tools
n8n community node	`@openserp/n8n-nodes-openserp`	Drop search, image, and extract steps into n8n automations

Both target Cloud and the self-hosted server with the same configuration — supply an API key for Cloud, or a base URL for a local server.

On the roadmap

Planned for upcoming releases:

Scheduled and recurring searches
Webhook or storage delivery for search jobs
AI extraction for summaries, entities, and structured fields
Web archive snapshots
Per-team keys and per-key usage breakdowns

What would unblock your use case? Tell us at feedback@openserp.org or via the dashboard feedback form — early feedback directly shapes what ships next.