How to Search Baidu Without a Chinese Phone Number

I wanted to check how a client's brand showed up on Baidu, and hit a wall fast. Baidu gates much of its search behind a Chinese phone number and a local account, and its anti-bot layer punishes anything that looks automated. I did not want to buy a SIM or rent a proxy farm just to read a results page, so I wrapped the whole thing in an Actor.

What this Baidu search API does

The Baidu Search Scraper is a Baidu search API: you send a query and it returns Baidu's organic results as structured JSON, one row per result. No Chinese phone number, no local account, and no browser to babysit.

What it returns

Every result comes back as a clean row with these fields:

• Title of the result

• Destination URL and display URL

• Snippet text

• Rank position on the page

You can pull a single page or paginate through several, then export straight to CSV or JSON. Device targeting, localization, and a time filter are built in.

Run it in three steps

1. Enter your query, and optionally set the device, region, and a time window.

2. Run the Actor and watch the results fill the dataset.

3. Download the rows as JSON or CSV, or read them from the API.

Use it as a REST API

Every run is reachable over REST, so you can trigger it from any language and read the dataset back as JSON. Send a request to the Actor run endpoint with a small JSON body:

POST /v2/acts/johnvc~Baidu-Search-Scraper/runs

{ "query": "electric vehicles", "maxPages": 2 }

Wire it into Claude with MCP

The Actor runs over MCP, so Claude or any MCP client can call it as a tool. Ask "what are the top Baidu results for electric vehicles in Beijing" and you get structured rows back, with no scraping code in your prompt.

What people build with it

People use it to track where a brand ranks on Baidu over time, to export search results into a spreadsheet for analysis, to feed a China-market research pipeline, and to run keyword rank checks from outside China without a local account.

Examples to copy

Four runnable examples ship on the Actor page, each with Chinese-language versions:

• Export Baidu Search Results to CSV

• Get Baidu Search Results in Claude via MCP

• Check How Your Brand Ranks on Baidu

• Search Baidu by API Without a Chinese Number

FAQ About Scraping Baidu

Where does the Baidu data come from?

The results are live Baidu organic listings, pulled at request time, so titles, links, snippets, and rank reflect what Baidu shows right now. Coverage is the organic results page, not paid ads or vertical widgets.

How do I call the Baidu scraper as an API?

Send a request to the Actor's run endpoint with your query, then read the dataset back as JSON or CSV. Billing is per page, so a single-page query costs a fraction of a cent and there is no monthly rental.

Can I use it from Claude with MCP?

Yes. The Actor is exposed over MCP, so you can add it to Claude or Cursor and ask for Baidu results in plain language. The client calls the tool and hands you structured rows.

Do I need a Chinese phone number or account?

No. That is the point of this scraper. You send a query and get results back without a Chinese SIM, a local login, or your own proxy setup.

If something looks off, open an issue on the Actor page and I usually reply within a day.