Weibo Scraper: Posts, Comments, Hot Search (微博热搜)

Weibo Scraper pulls public Sina Weibo (微博) posts by keyword search, user timeline or post URL, plus comments and the hot-search board, into JSON, CSV or Excel, with no login, cookies or API key. Each post comes with full text, an ISO 8601 timestamp, like / repost / comment counts, images, video, the poster's region and the author. It uses the anonymous guest session any first-time visitor gets, so there is no browser to run and no account to set up, and it costs $3 per 1,000 posts.

What you can do with it

• Brand and product monitoring. Track every mention of your brand, competitors or a product launch (瑞幸, iPhone 17, Nike) as it happens. Results come newest first.
• China social listening and market research. See what Chinese consumers say about a topic, with the poster's province and engagement numbers.
• Academic and media research. Collect public discourse on an event or hashtag for analysis.
• LLM and RAG pipelines. text is clean plain text, one field per post, ready for embedding, summarising or translation.
• Trend tracking. Turn on hotSearch for the current 微博热搜 board (rank, word, heat, label).
• Influencer scouting. Turn on includeUserProfile to get each author's follower count, bio and verification status.

Input

Field	Type	What it does
searchQueries	list of text	Keywords, Chinese or English. One search per keyword.
userIds	list of text	Numeric user IDs or profile URLs (https://weibo.com/u/1669879400, https://m.weibo.cn/u/1669879400). Returns the user's latest posts.
postUrls	list of text	Single posts: https://weibo.com/1785807165/RlqrBy6x0, https://m.weibo.cn/detail/5351065438128262 or a bare post ID. One request per post; deleted or private posts are skipped with a warning.
hotSearch	true / false	Default false. Adds the current 微博热搜 board, one row per trend (type is hotSearch: rank, word, heat, label, category, isTopic, url). Ads are skipped.
maxHotSearchItems	number	Trends to keep from the top of the board. Default 50, max 60.
maxItemsPerQuery	number	Cap per keyword and per user. Default 100, max 1000.
sinceDate	text	Optional. 2026-10-01 or full ISO 8601 (Beijing time if no timezone). Older posts are skipped; the run stops once a page holds nothing newer.
untilDate	text	Optional. Skips posts newer than this; a bare date means through the end of that day. For keyword search Weibo searches back from this moment, so together with sinceDate it reaches older date ranges, still up to 1,000 posts per keyword. sinceDate must be earlier than untilDate.
includeComments	true / false	Default false. Adds a comments list to each post: top-level comments, hottest first, with the first replies Weibo shows. Costs extra requests (about one per 20 comments).
maxCommentsPerPost	number	Top-level comments per post when comments are on. Default 20, max 1000.
includeUserProfile	true / false	Adds the exact follower count, bio, verified reason, following count, post count, gender and avatar to each author. One extra request per distinct author, so runs are slower.
requestDelaySeconds	number	Pause between requests, default 1.2 s (randomised up to 1.5x).
proxyConfiguration	object	Optional Apify Proxy. Default is no proxy.

Give at least one of searchQueries, userIds or postUrls, or turn on hotSearch. With none of them, the Actor runs a built-in example (keyword 咖啡, up to 5 posts from about a day ago) instead of failing.

{
  "searchQueries": ["咖啡", "luckin coffee"],
  "userIds": ["https://weibo.com/u/1669879400"],
  "maxItemsPerQuery": 200,
  "sinceDate": "2026-10-01",
  "untilDate": "2026-10-07",
  "hotSearch": true,
  "includeUserProfile": false
}

Output

One dataset item per post, plus one per trend when hotSearch is on. Export as JSON, CSV, Excel or via the API.

Field	Meaning
type	post for posts, hotSearch for hot-search rows (which carry rank, word, heat, label, category, isTopic, url and scrapedAt instead of the post fields).
id, mid	Weibo post ID (string).
url	Link to the post on weibo.com.
createdAt	ISO 8601 with timezone (Weibo shows Beijing time, +08:00).
text	Post text with HTML removed. Emoji appear as [name], line breaks as \n.
textRaw	Weibo's markup for the text with zero-width characters removed. On long posts it is the markup preview unless Weibo returns the full text as markup.
isLongText	true if Weibo had truncated the post; text then holds the full text.
repostsCount, commentsCount, likesCount	Engagement counts.
topics	Hashtags without the # marks, in order of appearance.
mentions	@ handles mentioned in the text.
pics	Image URLs, largest size available.
picCount	Number of images Weibo reports for the post.
video	{url, duration (seconds), playCount, cover} or null.
region	Poster's province / country as Weibo shows it ("发布于 江苏" becomes 江苏).
source	Client the post was sent from (for example iPhone客户端).
author	{id, screenName, profileUrl, followersCount, verified, verifiedReason, description}; with includeUserProfile also followingCount, postsCount, gender, avatarUrl.
retweetedStatus	The original post when this one is a repost, same shape; otherwise null.
comments	Only with includeComments: {id, postId, createdAt, text, likesCount, repliesCount, region, author, replies}.
query / userId / postUrl	The keyword, user or post URL the post was found through.
scrapedAt	UTC time of collection.

{
  "type": "post",
  "id": "5351532104777774",
  "mid": "5351532104777774",
  "url": "https://weibo.com/7564024690/RlCAik2US",
  "createdAt": "2026-10-07T23:49:02+08:00",
  "text": "#宋亚轩古茗咖啡全球代言人# ... 暮歌丨宋亚轩的微博视频 @时代少年团-宋亚轩",
  "textRaw": "<a href=\"//s.weibo.com/weibo?q=...\" target=\"_blank\">#宋亚轩古茗<span style=\"color: red;\">咖啡</span>全球代言人#</a> ...",
  "isLongText": false,
  "repostsCount": 0,
  "commentsCount": 0,
  "likesCount": 0,
  "topics": ["宋亚轩古茗咖啡全球代言人"],
  "mentions": ["时代少年团-宋亚轩"],
  "pics": [],
  "picCount": 0,
  "video": {"url": "http://f.video.weibocdn.com/o0/9faVhSlilx08zIWo7ciI01041203E6Ci0E020.mp4", "duration": 154.0, "playCount": 174000, "cover": "https://wx2.sinaimg.cn/orj480/008gImyHly1ifr8cel3rhj30u01hcju1.jpg"},
  "region": "安徽",
  "source": "Android客户端",
  "author": {"id": "7564024690", "screenName": "椿岛轩巷", "followersCount": null, "verified": true, "verifiedReason": null, "description": null, "profileUrl": "https://weibo.com/u/7564024690"},
  "retweetedStatus": null,
  "query": "咖啡",
  "scrapedAt": "2026-10-07T15:49:05+00:00"
}

Keyword search does not carry the author's follower count, bio or verified reason, so those are null unless you turn on includeUserProfile. Posts found through userIds always include them.

How much does it cost to scrape Weibo?

Pay per event: $3 per 1,000 posts (one post event, $0.003, for each post saved to the dataset). With includeComments on, each comment is a comment event at $2 per 1,000 ($0.002). With hotSearch on, each hot-search board entry is a trend event at $2 per 1,000 ($0.002). You pay only for posts you receive: nothing for failed pages or retries, and the run stops by itself when your maximum total charge is reached. Platform usage is included in the price.

Example: 5 keywords at 200 posts each is 1,000 posts, so $3. A daily brand-monitoring run of 100 new posts costs about $0.30.

Limits

• About 1,000 posts per keyword. Weibo serves roughly 25 pages of keyword search results (about 1,000 to 1,200 posts), however large the topic. To go deeper, split a topic into several narrower keywords or run it daily with sinceDate, or slice the period with sinceDate and untilDate.
• User timelines return the latest posts only (about 10). Without a login, Weibo shows guests the first page of a user's timeline and asks for a login on page 2. The Actor stops cleanly there. To follow a person or brand over time, run it on a schedule and dedupe on id, or search for their name or hashtag.
• Public data only, no login. Private, followers-only and age-restricted posts and follower lists are not available. A few posts are hidden from guests; they come back truncated rather than failing the run.
• Weibo can change or throttle its guest access at any time. The Actor refreshes its guest session, backs off and retries on errors, but a run may return fewer posts than requested when Weibo limits it. If that happens, raise requestDelaySeconds or enable a proxy.

Use with Python

# pip install apify-client
from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("datagleaner/weibo-scraper").call(
    run_input={"searchQueries": ["瑞幸"], "maxItemsPerQuery": 10}
)
for item in client.dataset(run.default_dataset_id).iterate_items():
    print(item["createdAt"], item["likesCount"], item["text"][:80])

Ten posts cost $0.03.

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/weibo-scraper').call({
    searchQueries: ['瑞幸'],
    maxItemsPerQuery: 10,
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
for (const post of items) console.log(post.createdAt, post.likesCount, post.text.slice(0, 80));

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

Actor ID: datagleaner/weibo-scraper

Minimal input:

{"searchQueries": ["瑞幸"], "maxItemsPerQuery": 10}

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/weibo-scraper 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/weibo-scraper. Without a token, an agent can still find Actors with the search-actors tool, but running this one needs your Apify account. In Claude Code:

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

  Then run /mcp to sign in to Apify in your browser. Then ask in plain language, for example: "Use the Weibo scraper to get the 50 newest Weibo posts mentioning 瑞幸 since 2026-10-01, and summarise in English what people complain about most, with the like counts."

• LangChain (Python):

# pip install langchain-apify, then set APIFY_TOKEN in your environment
import json
from langchain_apify import ApifyActorsTool
tool = ApifyActorsTool("datagleaner/weibo-scraper")
result = tool.invoke({"run_input": json.loads('{"searchQueries": ["瑞幸"], "maxItemsPerQuery": 10}')})

FAQ

How do I scrape Weibo posts by keyword? Put one or more keywords in searchQueries and set maxItemsPerQuery. Each keyword runs a Weibo search and returns the newest public posts first, up to about 1,000 per keyword. English keywords work, but Chinese keywords return far more.

Can I scrape a single post or go back to older dates? Put post links in postUrls. For older posts, set untilDate (with sinceDate) on a keyword search to reach earlier date ranges, up to 1,000 posts per range.

Can I scrape a Weibo user's posts? Yes, put a numeric user ID or profile link in userIds. Without a login Weibo shows guests only the latest page (about 10 posts); see Limits.

Do I need a Weibo API key, account or cookies? No. The Actor uses the anonymous guest session any visitor gets, so there is no Weibo developer app, login or cookie to set up. You only need an Apify account to run it.

Can I get Weibo hot search, comments or follower lists? Hot search and comments yes: set hotSearch for the 微博热搜 board and includeComments for top-level comments with their first replies (each is billed as described under pricing). Follower lists are not included.

Why did my run return fewer posts than maxItemsPerQuery? The topic has fewer public matches, sinceDate cut the list, Weibo's result cap was reached, or Weibo throttled the session. The run log says which; raising requestDelaySeconds or enabling a proxy helps with throttling.

Responsible use

This Actor collects only publicly visible posts. Respect Weibo's terms and keep the request rate low (the default pacing is deliberately polite). Posts and profiles are personal data: you are responsible for having a lawful basis and meeting your obligations under GDPR, China's PIPL and any other law that applies to your use, including retention, deletion requests and cross-border transfer. This is not legal advice.

中文说明

微博抓取器： 无需登录、无需 Cookie、无需浏览器，按关键词、用户或帖子链接抓取微博公开帖子，并可获取评论和微博热搜榜，输出干净的 JSON：正文（已去除 HTML）、ISO 8601 时间、转发/评论/点赞数、图片、视频、发布地区、来源、作者信息、被转发的原微博。

• 用途： 品牌监测、舆情与社交聆听、学术研究、喂给大模型的数据管线、达人筛选。
• 输入： searchQueries（关键词，中英文均可）、userIds（数字 UID 或主页链接）、maxItemsPerQuery（每个关键词/用户的上限，默认 100）、sinceDate / untilDate（起止日期）、postUrls（单条帖子）、includeComments / maxCommentsPerPost（评论）、hotSearch（热搜榜）、includeUserProfile（补全作者粉丝数、简介、认证信息）。
• 价格： 按条计费，帖子每 1,000 条 3 美元；开启评论时每 1,000 条评论 2 美元，开启热搜时每 1,000 条热搜 2 美元。只为实际保存的数据付费，达到你设置的最大花费后自动停止。
• 限制： 每个关键词最多约 1,000 条（微博搜索的上限）；游客状态下用户主页只能获取最新约 10 条；只抓取公开数据。
• 合规： 请遵守微博服务条款并保持较低的请求频率；个人信息的处理须符合 GDPR、《个人信息保护法》等适用法律，责任由使用者承担。