Baidu Web, Image & News Results Extractor

Baidu is the dominant search engine for Chinese-language content, and its result pages work differently from Google's — different ranking signals, different SERP layout, and separate verticals for web, image, and news queries. This actor runs one or more queries against Baidu's web search, image search, or news search and returns the results as structured rows, including Baidu's own search-operator filters (site, filetype, exact phrase, exclude words, title-only).

What this actor does

• Runs any number of queries against one of three Baidu verticals per run: web, image, or news.
• Applies Baidu's native search operators from simple input fields — no need to hand-build query strings for site:, intitle:, exact phrases, or excluded terms.
• Restricts web and news results to a publication window (last day, week, month, or year).
• Captures the "related searches" Baidu shows under organic web results as their own records.
• Resolves Baidu's redirect links to the real destination URL when the direct link isn't already present on the page.
• Fields a result doesn't have (e.g. no snippet, no image dimensions) are left out of that record.

Output per result (by mode)

Web search (mode: webSearch)

• query, position, page — the query text, rank position, and result page it came from
• resultType — organic for a normal result, related_search for a related-search suggestion
• title, url, snippet, date — standard result fields (present when Baidu shows them)
• displayUrl — the domain shown under the result title
• baiduUrl — Baidu's own redirect link, kept alongside url when one was present
• searchTerm — present only on related_search records; the suggested query text
• scrapedAt — collection timestamp

Image search (mode: imageSearch)

• query, position
• title — caption or source page title, when available
• imageUrl, thumbnailUrl — full-size and thumbnail image URLs
• sourceUrl, sourceDomain — the page the image was found on
• width, height — pixel dimensions, when Baidu reports them
• scrapedAt

News search (mode: newsSearch)

• query, position, page
• title, url, snippet, date
• source — the publisher/outlet name, when Baidu shows one
• scrapedAt

Input

Field	Type	Default	Description
queries	Array	["artificial intelligence"]	One or more search terms; each is scraped independently up to maxResults. Required.
mode	String (select)	webSearch	webSearch, imageSearch, or newsSearch.
maxResults	Integer	50	Maximum results to return per query, across all result pages (1-500).
timeRange	String (select)	any	any, day, week, month, year. Applies to webSearch and newsSearch only.
language	String (select)	any	any, simplified, traditional. webSearch only.
filetype	String (select)	any	any, pdf, doc, xls, ppt, rtf. webSearch only.
siteRestrict	String	–	Limit results to one domain, e.g. wikipedia.org. webSearch only.
exactPhrase	String	–	Require this exact phrase in results. webSearch only.
excludeWords	Array	[]	Words to exclude from results. webSearch only.
titleOnly	Boolean	false	Match the query against page titles only. webSearch only.
resolveUrls	Boolean	true	Follow Baidu's redirect link to the real destination URL in the rare case it wasn't already extracted directly from the page. webSearch only.
proxyConfiguration	Object	–	Apify Proxy settings; recommended when requesting from outside China.

Example: Simplified Chinese web results from the last month

{
  "mode": "webSearch",
  "queries": ["人工智能 最新进展"],
  "maxResults": 50,
  "timeRange": "month",
  "language": "simplified"
}

Example: image search

{
  "mode": "imageSearch",
  "queries": ["Great Wall of China"],
  "maxResults": 100
}

Example: news search restricted to the last week

{
  "mode": "newsSearch",
  "queries": ["technology innovation"],
  "maxResults": 30,
  "timeRange": "week"
}

Example: PDF files on one domain, excluding a term

{
  "mode": "webSearch",
  "queries": ["年度报告"],
  "filetype": "pdf",
  "siteRestrict": "gov.cn",
  "excludeWords": ["draft"]
}

Use cases

• Baidu SEO tracking — check where your pages rank for target keywords on Baidu over time.
• Competitor monitoring — see which domains and content types rank for a competitor's brand or product terms.
• Chinese media monitoring — pull news coverage of a topic, brand, or event from Baidu News.
• Academic and research sourcing — collect Chinese-language sources on a topic using filetype and siteRestrict filters.
• Visual research — gather image results and their source pages for a topic with imageSearch.

FAQ

Why would a run return zero results? Baidu is more likely to throttle or block requests coming from outside China. Setting proxyConfiguration with Apify Proxy enabled generally resolves this.

What's a related_search record? It's the "related searches" suggestions Baidu shows beneath organic web results, captured with resultType: "related_search" and the suggestion text in searchTerm, rather than a regular result.

Does language translate my query? No — it restricts which script (Simplified or Traditional Chinese) the returned pages are written in; it doesn't translate or rewrite your search terms.

Why do some web results have a baiduUrl but others don't? Baidu usually embeds the real destination URL directly in the page, so url is already correct and baiduUrl is omitted. baiduUrl only appears when the page only exposed a Baidu redirect link for that result.

Can I combine exactPhrase, siteRestrict, and excludeWords in one query? Yes, these web-search operators combine freely in the same query.

Is there a limit on how many results I can get per query? Yes, maxResults caps out at 500 per query. Run the same query again or split it across multiple queries in one run if you need more coverage.

Does image search work the same as Google Images for the same query? No — Baidu's image index and ranking are independent of Google's, so results for the same query can differ noticeably.

Is this affiliated with Baidu? No. This actor reads Baidu's public search result pages and is not affiliated with Baidu, Inc.