WeChat Official Account Scraper: 公众号 articles as text and Markdown

Extract public WeChat Official Account (微信公众号) articles from mp.weixin.qq.com links, or find them by keyword or account name, and get title, account, publish time, full text, Markdown and images. No WeChat login, no cookies, no browser.

Price: $5 per 1,000 articles ($0.005 per article, full text included). 10 articles cost 5 cents. Failed or deleted articles are not charged.

Why use this one

• No login and nothing of yours at risk. It reads public article pages only. No WeChat account, QR login or cookies.
• Clean output for LLM and RAG pipelines. Plain text and Markdown (images kept as links), plus an optional HTML body, the 原创 original badge and the IP region shown on the article.
• Three ways in. Paste article links (most reliable), search by keyword (Chinese works best), or search by account name.
• Runs through Apify Proxy by default, because keyword search showed a block on a datacenter IP without it.
• Honest limits: keyword and account discovery relies on Sogou's public WeChat search (about 100 recent articles per query, best effort for account names); read and like counts are not public, so they are not in the output.

Minimal input

{"searchKeywords": ["新能源汽车"], "maxArticles": 20, "publishedWithin": "week"}

Sample output row (real platform run, long text shortened)

{
  "articleId": "Njk5MTE1_2652701954_1",
  "title": "杜绝“新马路杀手”，他们为新能源汽车设计声音",
  "accountName": "南方周末",
  "accountId": "nanfangzhoumo",
  "biz": "Njk5MTE1",
  "author": "林子人",
  "publishedAt": "2026-09-29T09:35:43+00:00",
  "publishTimestamp": 1790674543,
  "isOriginal": true,
  "ipRegion": "广东",
  "mid": 2652701954,
  "idx": 1,
  "digest": "▲ 广州南沙,黄阁汽车货运港码头的新能源汽车.(视觉中国/图)全文共5866字,阅读大约需要12分钟他觉得汽车的声音设计只是一...",
  "coverImage": "https://mmbiz.qpic.cn/mmbiz_jpg/Q3auHgzwzM54CUiaNdz8kKdhCZgib0FgqMbjngbb1bfwBicNVIUNQqxGKBW5oh4Avdnf7B7zEIhg3i...",
  "images": [
    "https://mmbiz.qpic.cn/mmbiz_gif/Tjnia6K0WAwwkPLicCXIlLH72QrQ0aaK9yTAGAZgAicrFQxtk5XBtibwZg1MV0icCj4F3heQvRiawN...",
    "https://mmbiz.qpic.cn/sz_mmbiz_jpg/Q3auHgzwzM4LAiaDy0vOUTJE9SC160YuqlogqBphyhtxVYKcJDZVeBCibYjzjXRcsMyclibprSg...",
    "https://mmbiz.qpic.cn/sz_mmbiz_jpg/Q3auHgzwzM6xS65Mas2qwXlVmtBmobM2Klr1AibqJibnsKCWOXhuXIfMuJeI2mmE1DPZC7gXjZu..."
  ],
  "contentText": "▲ 广州南沙，黄阁汽车货运港码头的新能源汽车。（视觉中国/图）\n\n全文共5866字，阅读大约需要12分钟\n\n他觉得汽车的声音设计只是一份谋生的工作而已，从未想过有朝一日城市的街道上奔跑的几乎都是新能源汽车，届时这些车辆发...",
  "contentMarkdown": "![图片](https://mmbiz.qpic.cn/mmbiz_gif/Tjnia6K0WAwwkPLicCXIlLH72QrQ0aaK9yTAGAZgAicrFQxtk5XBtibwZg1MV0icCj4F3heQ...",
  "url": "https://mp.weixin.qq.com/s?src=11&timestamp=1791700648&ver=7019&signature=DGUqn4QoyCG*0CHhG*uR2HGOyOM5F8-Iqq07...",
  "sourceType": "keyword",
  "sourceUrl": null,
  "sourceKeyword": "新能源汽车",
  "scrapedAt": "2026-10-11T06:38:04+00:00"
}

What you get

For each article: title, account name, account ID (gh_...), biz, mid and idx, a stable articleId, author byline, publish time (publishedAt ISO 8601 and publishTimestamp Unix seconds), isOriginal (the 原创 badge; null if the page does not say), ipRegion (the IP属地 shown under the title, null on articles from before 2022), digest, cover image, every image URL, and the body as plain text, Markdown and/or HTML. The dataset has two views: Articles (a table of metadata) and Content (full Markdown text, one row per article).

Use cases

• Build a Chinese-language news or industry-monitoring feed (AI, EV, retail, finance).
• Feed Chinese articles to an LLM or RAG pipeline as clean Markdown.
• Export or download articles as Markdown before they are deleted (WeChat removes content often).
• Track what specific Official Accounts publish.

Input

Field	Meaning
articleUrls	Public article links. The most reliable mode.
searchKeywords	Keywords to discover articles (via Sogou WeChat search). Chinese works best.
accountNames	Account display names. Best effort: searches the name and keeps results whose account name contains it (case-insensitive).
maxArticles	Cap per keyword or account (default 20, max 100). The Console prefill is 5.
publishedWithin	day, week or month (default: any time). Filters keyword and account results after search, so a narrow window may return fewer than maxArticles; results with no date are kept. Does not apply to articleUrls.
includeContent	text, markdown, html (default text + markdown).
requestDelaySeconds	Base pause between requests (default 1.5).
proxyConfiguration	Default: Apify Proxy (datacenter).

{
  "articleUrls": ["https://mp.weixin.qq.com/s/_2kC-fXw7UjneZSrsC9CVQ"],
  "searchKeywords": ["新能源汽车"],
  "accountNames": ["新华社"],
  "maxArticles": 20,
  "publishedWithin": "week",
  "includeContent": ["text", "markdown"]
}

An empty input runs a 5-article example for the keyword 人工智能, and the Console prefill also runs 5 articles.

Output

{
  "articleId": "MzY5ODQ1Njg1Ng==_2247483700_1",
  "url": "https://mp.weixin.qq.com/s?src=11&timestamp=1791388858&ver=7012&signature=...",
  "title": "谷歌签下史上最大核电协议，AI的尽头是电",
  "accountName": "海风蓝的专属小站",
  "accountId": "gh_b761fc45617b",
  "biz": "MzY5ODQ1Njg1Ng==",
  "author": null,
  "publishedAt": "2026-10-07T13:33:21+00:00",
  "publishTimestamp": 1791380001,
  "isOriginal": true,
  "ipRegion": "广东",
  "mid": 2247483700,
  "idx": 1,
  "digest": "你以为谷歌最近在忙什么？发布新模型？收购公司？……",
  "coverImage": "https://mmbiz.qpic.cn/...",
  "images": ["https://mmbiz.qpic.cn/..."],
  "contentText": "...",
  "contentMarkdown": "...",
  "sourceType": "keyword",
  "sourceKeyword": "人工智能",
  "sourceUrl": null,
  "scrapedAt": "2026-10-07T16:01:10+00:00"
}

For articleUrls items, sourceType is articleUrl and sourceUrl is your input link; for discovered items it is keyword or account, with the query in sourceKeyword. canonicalUrl appears when the page exposes a permanent link, and url then uses it. contentHtml appears when you ask for html. articleId (biz_mid_idx) is stable across runs; use it, not url, to de-duplicate or monitor, because signed search-result links expire.

Pricing

Pay per event: $5 per 1,000 articles (event article, one per article pushed). Example: 300 articles from three keywords cost about $1.50. There is no free-run cap. The run stops pushing once your max total charge is reached. Failed or deleted articles are not charged.

Limits

• Keyword and account discovery uses Sogou's public WeChat search: at most 10 pages (about 100 articles) per query, recent articles only, and it can show a verification page if hit too fast. The Actor paces itself, retries with a fresh session and, if still blocked, ends discovery for that query with a message. Keep the default Apify Proxy setting for keyword and account search; without a proxy, Sogou blocked a datacenter IP after about 12 articles in testing. Article URLs are not affected.
• When a page exposes no permanent link, an article found through search keeps the signed url from the search redirect, which WeChat expires after a while; use articleId to identify it long-term.
• Account lookup is best effort: Sogou's account directory is not available anonymously, so only articles surfaced by searching the name are returned.
• Deleted, reported or restricted articles, and some pure video or image-share posts, have no body and are skipped.
• Read and like counts are not public to anonymous visitors, so they are not in the output.

Use with Python

# pip install apify-client
from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("datagleaner/wechat-articles").call(run_input={
    "searchKeywords": ["新能源汽车"],
    "maxArticles": 5,
    "includeContent": ["markdown"],
})
for item in client.dataset(run.default_dataset_id).iterate_items():
    print(item["publishedAt"], item["accountName"], item["title"])

Five articles cost about $0.025.

Use with JavaScript / Node.js

// npm install apify-client
import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('datagleaner/wechat-articles').call({
    searchKeywords: ['新能源汽车'],
    maxArticles: 5,
    includeContent: ['markdown'],
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
for (const item of items) console.log(item.publishedAt, item.accountName, item.title);

Use it from n8n, Make, Zapier or an AI agent

Actor ID: datagleaner/wechat-articles

Minimal input:

{"searchKeywords": ["新能源汽车"], "maxArticles": 5}

Each tool below runs this Actor with your own Apify API token.

• n8n: add the Apify node (@apify/n8n-nodes-apify). On n8n Cloud you install it from the community node registry. Choose Run an Actor and get dataset, set Actor to datagleaner/wechat-articles and paste the input above.

• Make: use the Apify app's Run an Actor module, then Get Dataset Items to read the results. Watch Actor Runs can trigger a scenario when a run finishes.

• Zapier: use the Apify action Run Actor, then the search Fetch dataset items. The trigger Finished Actor run starts a Zap when a run ends.

• AI agents (MCP): connect to https://mcp.apify.com?tools=datagleaner/wechat-articles. In Claude Code:

  claude mcp add --transport http apify "https://mcp.apify.com?tools=datagleaner/wechat-articles"
  

  Then run /mcp to sign in to Apify in your browser. Other clients can send the header Authorization: Bearer YOUR_APIFY_TOKEN instead. Without a token the server only offers search and docs tools, so running the Actor needs one of the two.

• LangChain (Python):

# pip install langchain-apify, then set APIFY_TOKEN in your environment
import json
from langchain_apify import ApifyActorsTool
tool = ApifyActorsTool("datagleaner/wechat-articles")
result = tool.invoke({"run_input": json.loads('{"searchKeywords": ["新能源汽车"], "maxArticles": 5}')})

Claude Desktop and other clients that use an mcpServers config can run Apify's local server instead; the agent finds this Actor with the search-actors tool and runs it with call-actor:

{
  "mcpServers": {
    "actors-mcp-server": {
      "command": "npx",
      "args": ["-y", "@apify/actors-mcp-server"],
      "env": { "APIFY_TOKEN": "YOUR_APIFY_TOKEN" }
    }
  }
}

Example prompt:

Find the 10 most recent WeChat articles about 新能源汽车, then give me each one's title, account and publish date, and a three-sentence English summary of the Markdown body.

FAQ

How do I export a WeChat article to Markdown? Paste its mp.weixin.qq.com link into articleUrls and keep markdown in includeContent. Each article comes back with contentMarkdown plus every image URL, and you can download the dataset as JSON, CSV or Excel.

Do I need a WeChat account or login? No. Only public pages are fetched, with no login and no cookies.

How do I search WeChat Official Account articles by keyword? Put keywords in searchKeywords. Discovery runs through Sogou's public WeChat search, which returns recent articles only, at most about 100 per keyword.

Can it download every article of one Official Account? No. There is no public account archive, so accountNames keeps the search results published by that account, and a quiet account may return few. Supply links for full coverage.

Does it return read and like counts? No. WeChat does not show them to anonymous visitors, so the Actor does not output them.

Responsible use

This Actor reads publicly available pages only. You are responsible for using the data lawfully, including copyright and any personal-data obligations (for example GDPR or PIPL) that apply to what you collect and store.