WeChat Articles Scraper

WeChat Articles Scraper searches WeChat Official Account articles (公众号文章) by keyword, with no WeChat account, login or cookies, and returns the complete article in the same run: title, account, author, publish date, publishing region, original badge, digest, cover, images, hashtags, the full text and, on request, the HTML. It also pulls the articles of given Official Accounts and reads mp.weixin.qq.com links you already have. For China market researchers, brand and media monitoring teams, Chinese-language corpora and LLM datasets, newsrooms and AI agents.

At a glance

Apify plan	Article found	Full text	Article with full text, per 1,000
Free	US$ 0.0040	US$ 0.0100	US$ 14.00
Bronze	US$ 0.0038	US$ 0.0095	US$ 13.30
Silver	US$ 0.0036	US$ 0.0090	US$ 12.60
Gold and above	US$ 0.0034	US$ 0.0085	US$ 11.90

With includeContent: false you pay only the "article found" price (US$ 4.00 per 1,000 on Free). Links you send always include the full text. No monthly fee, no start fee, residential fallback included. One article is charged once per run however many keywords found it. Free: error rows, duplicates, articles outside your date filter and articles skipped by onlyNew. The default run (20 articles with full text) costs US$ 0.28.

What it fixes

• Search links expire; the text does not. WeChat search hands out signed links that stop working after a while and do not reveal the permanent address without a login. The Actor reads the full article in the same run, and gives articleKey plus accountBiz, messageId and indexInMessage as the permanent identity.
• Account data on every full text: the WeChat ID (微信号), gh_ id and the account description.
• Fields the search page hides: author, publishing region (发表于), the 原创 original badge and #话题 hashtags.
• Monitoring without paying twice. onlyNew remembers what each keyword already delivered and skips it before any charge.

Input examples

Every field is optional; the input form describes each one. An empty input searches 人工智能 (artificial intelligence) and returns 20 articles with the full text.

Watch three topics every morning (save as a Task with a daily Schedule):

{
  "queries": ["新能源汽车", "跨境电商", "人工智能"],
  "maxResultsPerQuery": 100,
  "onlyNew": true,
  "publishTime": "week"
}

The newest articles about a brand, search data only:

{ "queries": ["小米 汽车"], "sortBy": "latest", "maxResultsPerQuery": 50, "includeContent": false }

The articles of an Official Account, by name or by any article link from it:

{ "accounts": ["36氪"], "queries": [] }

Articles you already have, with HTML:

{ "articleUrls": ["https://mp.weixin.qq.com/s/XXXXXXXXXXXXXXXXXXXXXX"], "includeHtml": true }

With onlyNew, the first run delivers everything up to maxResultsPerQuery; later runs deliver only articles not seen before, and a quiet day ends with a free NOTHING_NEW row. The memory is the Key-Value Store wechat-articles-only-new in your account, one record per keyword. In a test on 2026-10-01, a 100 article seed of 人工智能 followed by a second run repeated none of the 100.

Switching from another WeChat actor? Your existing input works

Your field	Read as
keywords	queries
links	articleUrls
maxPages	maxResultsPerQuery (10 articles per page)
publishTime: all, halfYear	any time, the last 6 months
sortBy: "hot"	relevance
operation: articleSearch	keyword search
operation: articleDetail	articleUrls
operation: accountHistoryArticles, accountTodayArticles	accounts

Operations that need a logged in WeChat app (read and like counts, comments, WeChat Index, suggestions, mini programs, account owner) return one free row with errorCode: "UNSUPPORTED_OPERATION".

Output example

A real row from a run on 2026-10-01 with {"queries": ["人工智能"], "maxResultsPerQuery": 10}, long fields trimmed:

{
  "type": "article",
  "searchQuery": "人工智能",
  "position": 1,
  "page": 1,
  "articleKey": "7ec3b116eb797951",
  "title": "10月首批!重庆AI智能体培训启动,现已开始,助力冲刺大厂高薪offer,20岁以上可学",
  "summary": "2026年AI智能体培训启动为落实《关于深入实施“人工智能+”行动的意见》部署...",
  "accountName": "重庆潮生活",
  "accountId": "gh_31d5f2253ef7",
  "accountBiz": "MjM5ODMwMzExNQ==",
  "author": "AI智能体6",
  "publishedAt": "2026-10-01T03:03:41Z",
  "ipRegion": "广东",
  "isOriginal": false,
  "hashtags": [],
  "coverImage": "https://mmbiz.qpic.cn/mmbiz_jpg/smuH1LUwDYC8NyL0Xw1Cudv6lYSrsoIGTlicjFYrlO73A...",
  "content": "2026年AI智能体培训启动\n为落实《关于深入实施“人工智能+”行动的意见》部署...",
  "contentLength": 886,
  "images": [],
  "imageCount": 0,
  "messageId": "2652307344",
  "indexInMessage": "2",
  "articleUrl": "https://mp.weixin.qq.com/s?src=11&timestamp=1790862968&ver=7000&signature=G5lkXYyuSBz1q7ro...",
  "scrapedAt": "2026-10-01T13:56:08Z",
  "errorCode": null,
  "error": null
}

Main output fields

Field	What it is
searchQuery, position, page	The keyword and the rank on WeChat search
articleKey	Stable id across runs; use it to deduplicate your storage
title, summary, accountName, publishedAt, coverImage	What the search page shows
articleUrl	The link the article was read from (from a search, a signed link that expires)
digest, author, accountId, accountBiz, accountAvatar	Article digest, signed author, the account's gh_ id, __biz and avatar
accountWechatId, accountDescription	The account's WeChat ID (微信号) and self description
accountQuery	The account you asked for, on rows from accounts
content, contentLength, contentHtml	Full body as plain text (one paragraph per line), its length, and HTML with includeHtml
images, imageCount	Every image in the body
isOriginal, ipRegion, hashtags	原创 badge, region under 发表于, #话题 tags
messageId, indexInMessage	With accountBiz, the permanent identity of the article
postType	image_post for photo posts (小绿书)
contentError	The body could not be read; search data delivered and only the search price charged

The Console shows two views: Overview and Full text.

Errors (never charged)

errorCode	Meaning
NO_RESULTS	No article for this keyword in the chosen period
NOTHING_NEW	With onlyNew, everything listed was already delivered
LINK_NOT_RESOLVED	The search did not release the article's link; title, account, date and digest still delivered
BLOCKED	The search page did not come back after several fresh sessions; run again in a few minutes
NOT_REACHED	The run timeout arrived before this keyword
NOT_FOUND, DELETED, VERIFICATION	A link whose article is deleted, private or behind verification
INVALID_URL, INVALID_INPUT	Not an mp.weixin.qq.com link, or a filter value that is not valid (a gh_ id alone cannot be searched)
UNSUPPORTED_OPERATION	An operation that needs a logged in WeChat app

Use it from code and AI agents

from apify_client import ApifyClient

client = ApifyClient("YOUR_TOKEN")
run = client.actor("scrapewise/wechat-articles").call(run_input={"queries": ["跨境电商"], "sortBy": "latest"})
for row in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(row["publishedAt"], row["accountName"], row["title"])

The same works from Node.js, the HTTP API (run-sync-get-dataset-items), n8n, Make and Zapier. For AI agents, add scrapewise/wechat-articles through the Apify MCP server (https://mcp.apify.com).

Limitations

• About 100 articles per keyword. WeChat search shows 10 pages to a logged out visitor, and this Actor never logs in. sortBy, publishTime and publishedAfter order and filter those 100; they do not dig deeper. Use several keywords and onlyNew day after day for more coverage.
• No read counts, likes, comments or follower numbers: those need a logged in WeChat client.
• An account's most visible articles, not its whole archive. accounts searches the name and keeps only that account's articles (24 of 26 results for 36氪 on 2026-10-08). A very generic name can return none, free.
• Some articles cannot be read (deleted, restricted, verification); they come with contentError at the search price only.
• Content is in Chinese and not translated. Chinese keywords return far more than English ones.

FAQ

What should I store, since the link expires? The full text, articleKey, and accountBiz + messageId + indexInMessage.

How do I get more than 100 articles on a topic? Several keywords in one run (duplicates are free) and a daily onlyNew schedule.

Is there any ban risk? No WeChat account, session or cookie is used.

Something broke? Open an issue on the Issues tab.

Changelog

• 2026-10-08: accounts for the articles of given Official Accounts; accountWechatId and accountDescription on every full text; the input of other WeChat actors accepted as it is.
• 2026-10-01: onlyNew, sortBy: latest, publishTime; isOriginal, ipRegion, hashtags, articleKey and includeHtml.
• 2026-09-29: links resolved in parallel, about twice as fast; resumes after a platform restart without charging twice.