China Social Media Search - Weibo, WeChat, Douyin & More

Search Chinese social media and news with one keyword, without a login. One run covers Weibo (微博), WeChat official accounts (公众号), Zhihu (知乎), Douyin (抖音), Xiaohongshu (小红书, RedNote), Bilibili (B站), Kuaishou (快手) and Chinese news sites, inside a time window you choose. Built for China brand monitoring, market research, PR and competitor tracking.

What is China Social Media Search?

Most Chinese platforms cannot be searched from outside China, and each one needs its own tool. This Actor searches them together and returns one flat list of posts with the same field names on every platform. It is not affiliated with any of the platforms; it reads what anyone can see in their public search.

What data can you get?

One record per post (type: post):

Field	Meaning
platform	news, weibo, wechat, zhihu, douyin, xiaohongshu, bilibili or kuaishou; other for a site that is none of them (for example a video site). In a search over all platforms it is worked out from the site name, so check siteName when it matters
siteName	The site or account type as the source names it, for example 微信公众号, 新浪微博, 网易号
keyword	The keyword that found the post
title, text	Title (empty for short posts such as Weibo or Douyin) and the text or summary returned by the search; articles can be long
url	Link to the post
authorName, authorId, authorAvatarUrl	The author as the post shows it publicly (news sites often have none)
publishedAt	Publication time, ISO 8601, UTC. Some sites give only the date; it then shows as midnight Beijing time (16:00 UTC of the day before)
imageUrls, videoUrl	Picture and video addresses of the post; they can be temporary

The search does not return likes, comments or view counts. For those, use the single-platform scrapers listed at the end.

How to use China Social Media Search

1. Enter one or more keywords (Chinese works best).
2. Optionally pick platforms and a time window.
3. Set the maximum number of results per search and run.
4. Export the dataset as JSON, CSV or Excel, or read it through the API. Schedule the Actor to monitor a brand every hour or day.

Input

Field	Type	Meaning
keywords	list	Search queries. Each keyword is a separate search and pays its own start fee. Falls back to 小米 when empty
platforms	list	Limit the search to some platforms. Empty (or all eight) searches every platform in one search. Each keyword and chosen platform is a separate search and pays its own start fee
lastHours	integer	Only posts of the last N hours, 1 to 744. Default 24
startTime, endTime	string	Optional window in UTC, such as 2026-10-21 or 2026-10-21 12:30. startTime overrides lastHours; endTime defaults to now. The window is at most 744 hours (31 days)
maxResults	integer	Posts per search, up to 1000. Default 20

Query syntax

A keyword can combine terms:

Syntax	Meaning
A && B	contains both A and B
A || B	contains A or B
-A	does not contain A
(A && B) || C	grouping
"A"	exact phrase, no word splitting

Example, coffee brand news without job ads, on three platforms:

{"keywords": ["\"瑞幸咖啡\" && (新品 || 联名) -招聘"], "platforms": ["weibo", "wechat", "xiaohongshu"], "lastHours": 72, "maxResults": 40}

Example, one keyword on all platforms over a fixed day:

{"keywords": ["小米"], "startTime": "2026-10-21", "endTime": "2026-10-22", "maxResults": 50}

Output example

{"type": "post", "platform": "wechat", "siteName": "微信公众号", "keyword": "小米", "title": "小米澎程N90参数很猛，真正要核实的是身份",
 "text": "...", "url": "https://mp.weixin.qq.com/s?__biz=MzY4NjQwODc1Mw==&mid=2247485948&idx=3", "authorName": "初心飘香4056",
 "authorId": "gh_59f4c54ecaf6", "publishedAt": "2026-10-14T16:00:00Z", "imageUrls": ["https://mmbiz.qpic.cn/..."], "videoUrl": null}

How much does it cost?

You pay per event: a start fee of $0.04 per search, plus $0.02 per post returned.

• Every further keyword pays the start fee again, and so does every further keyword and platform pair when you choose platforms.
• A post already returned by an earlier search of the same run is not returned or charged twice.
• Example: 40 posts for one keyword cost $0.04 + 40 x $0.02 = $0.84.
• Example: 3 keywords with 20 posts each, platforms left empty (all eight in one search each), cost 3 x $0.04 + 60 x $0.02 = $1.32.
• Example: 1 keyword on 3 chosen platforms with 40 posts each is 3 searches and costs up to 3 x $0.04 + 120 x $0.02 = $2.52.
• Each keyword and platform pair is a separate search with its own start fee. Leaving platforms empty searches all eight platforms in one search with one start fee.
• A spending limit set on the run is respected: the run stops cleanly when the next post or start fee does not fit.

Free Apify plan: every run returns a sample dataset (a real result of the prefilled input) so you can see the output format; your input is not used and nothing is charged. To run your own queries, use any paid Apify plan.

Use cases

• Watch what Chinese social media and news say about your brand or product, every hour or every day.
• Compare how a topic is covered on Weibo, WeChat, Xiaohongshu and Douyin.
• Find the articles, videos and posts behind a trend, a launch or a crisis in a chosen period.
• Track competitors by name, with exclusion terms to cut recruitment and advertising noise.

Integrations and API

Runs can be started through the Apify API, scheduled, and connected to integrations such as webhooks, Zapier or Make.

from apify_client import ApifyClient
client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("clover_folktale/china-search-scraper").call(run_input={"keywords": ["小米"], "platforms": ["weibo"], "lastHours": 6})
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["platform"], item["publishedAt"], item["title"] or item["text"][:60], item["url"])

Tips and limits

• Posts come oldest first inside the time window. A long window on a popular keyword fills your quota with the oldest posts, so use a short window (a few hours) and schedule the Actor to get the newest ones.
• A keyword matches any post that contains it, so a short or common word also brings unrelated posts (searching 小米 alone returns posts that only mention it in a hashtag). Narrow it with the query syntax, for example "小米" && (手机 || 汽车).
• A page holds 10 posts. The search goes on while a page brings at least 2 posts you do not have yet and the platform has more; it stops at your maxResults.
• Times are UTC. The platforms publish in China time (UTC+8), so a post made at 08:00 in Beijing has publishedAt 00:00Z.
• A quiet keyword or a short window can return nothing for a platform; that is a normal result, and the start fee still applies because the search was made.
• Platforms and news sites differ in how much text they return; Weibo and Douyin posts have no title.
• Only what the post shows publicly is returned: text, public author name and ID, picture and video addresses. No private data.
• Picture and video addresses can be temporary; download what you need soon after the run.

FAQ

Do I need a Chinese account or a login? No login or cookies are needed.

Which platforms are covered? News sites, Weibo, WeChat official accounts, Zhihu, Douyin, Xiaohongshu, Bilibili and Kuaishou.

Is it legal? Only public data is returned. No messages, phone numbers or other private data. You are responsible for how you use it.

How fresh is the data? Each run reads the platforms at that moment; posts appear a few minutes to hours after publication.

Can I get comments, likes or follower counts? Not from this Actor. Use the single-platform scrapers below.

Can I schedule runs and export? Yes, to JSON, CSV, Excel and more.

More tools from the same developer

• Weibo Scraper - Hot Search, Posts, Comments & Users
• WeChat Official Account Scraper - Articles & Read Counts
• Xiaohongshu RedNote Scraper - Notes, Comments & Replies
• Douyin Scraper - Videos, Comments, Profiles & Hot Search
• Zhihu Scraper - Answers, Articles & Comments
• Bilibili Scraper - Videos, Comments, Danmaku & Subtitles