Skip to main content
Octoparse Data Hub turns a data capability into something software can call. Scraping templates, scraping scripts, existing databases, third-party APIs, and processing programs all get wrapped in one shape: the Data App. A person can run one in the browser. An agent can find one, read its contract, and call it without anyone writing glue code.

Open Octoparse Data Hub

See what is live today and filter by capability, industry, platform, and content type.

​
Why Data Hub

Data products used to be built for people to read. You opened a site, downloaded a report, and cleaned up the result by hand. An agent cannot do any of that. It needs something it can call with structured arguments, a description it can parse on its own, and a price it can be billed for by usage. That is the gap Data Hub fills. You stop shopping for a vendor per source, writing a scraper per site, and adapting to a new interface per API. Every capability is discovered and called the same way, whether a person or an agent is driving it.

​
How Data Hub works

Scraping templates, scripts, datasets, third-party APIs, and processing programs are wrapped as Data Apps, then served to agents, business systems, and people through the web portal, API, MCP, or SDK The source underneath can be a scraper, a dataset you already own, a partner API, or a cleaning routine. Callers never see the difference. That is the point: the catalog can keep growing while the integration you wrote last quarter keeps working.

​
What is a Data App

A Data App is one callable capability. Enrich a company name with firmographics, pull reviews for a product URL, fetch the details behind a video link, normalize a block of messy text. Each of those is packaged, priced, and versioned on its own. A Data App carries a publisher and app name, a machine-readable spec, input and output fields, a public price, and a run-it-now entry point

​
Who uses it

Teams that consume data

Agent and application builders, analysts, growth and e-commerce teams, and engineers who need outside data inside their own systems. Find a capability, try it in the browser, wire it into whatever you already run.

Teams that publish data

The Octoparse data team, plus companies and developers sitting on data assets, collection know-how, or processing pipelines. Wrap a capability as a Data App, set a price, and serve callers under platform rules.
Publishing review, visibility controls, content governance, and category upkeep keep the catalog worth searching.

​
What people build with it

Lead enrichment

Hand your agent a list of company names and get industry, headcount, and headquarters back, straight into the CRM record or the outbound sequence.

Competitor price tracking

Watch listings, promotions, and stock status on a schedule, then feed a pricing dashboard or a repricing rule.

Search visibility

Pull ranked results for the queries you care about and track how position, snippets, and competitors move over time.

Review and reputation analysis

Collect reviews across marketplaces and local listings, then run sentiment and theme extraction on top of them.

Talent and hiring research

Track public job postings to read demand by role, seniority, and region before you plan headcount or pricing.

Tools for your own agent

Register the apps your workflow needs, and let the agent decide when to call them mid-conversation or mid-task.

​
Four ways in

Data Hub is reachable four ways: the web portal, the REST API, MCP, and the official SDKs

Web portal

Where evaluation starts. Browse, read the contract, run it, and check or export the result in the console.

API

For teams with engineers. An API key and standard endpoints put records straight into your systems and pipelines.

MCP

For anyone working through an agent. Connect once and let it search, or pin a single app as a fixed tool.

SDK

For fast integration. The official packages handle auth, polling, and paging so you write the business logic only.
Same account, same catalog, same results behind all four. Pick the one that matches how you already work.

Data Hub MCP capabilities

The six tools of the general MCP server, synchronous versus asynchronous runs, result paging, and how oversized results get handed off for export.

​
What it costs

You browse and read for free, and pay only when a call actually returns data. Each publisher sets the rules, published on the app’s detail page. Three patterns cover almost everything:
  • Base fee: a flat charge per submitted run, independent of how many records come back.
  • Per record: charged on records returned successfully, sometimes bundled into blocks of a fixed size.
  • Custom unit: whatever unit fits the capability, such as a minute of audio.
A failed run costs nothing. A partial run bills only the records it delivered. Charges come out of your account balance, and new calls stop when the balance runs out. Usage and invoices live in the Open Platform console.

​
Getting started

1

Pin down what you need

The business question, the sources, the fields, and how fresh the data has to be.
2

Find the app

Search Octoparse Data Hub and read the capability description, the arguments, the returned fields, and the price.
3

Run it once

Sign in, fill in the arguments, and run. Confirm the records look like what you expected before writing any code.
4

Wire it in

Move downstream through the portal, the API, MCP, or an SDK, whichever fits your stack.
Data Hub is a catalog of callable data capabilities. The Octoparse MCP Server is a different product that runs scraping templates and cloud tasks on your own account, documented under MCP Server. To try an app right now, read Run a Data App in the browser. If you have not picked one yet, start with General connection. If you already know which app you want, go to Codex or Claude Code.