Weibo Scraper

A Weibo scraper for Sina Weibo (微博): keyword search, a user's posts, fans and following, comments, the hot search list (热搜) and the hot feed. No browser.

$4.49 per 1,000 items, with or without your own cookie.

Quick start

Runs as-is with this input (the same input Apify tests daily):

{
  "searchKeywords": [
    "新能源汽车"
  ]
}

Main output fields: type, createdAt, text, likes, comments, reposts, region, keyword, url, word, heat, name, followers.

Price: Result: $4.49 per 1,000; Hot search change row: $25 per 1,000. Pay only for results delivered.

Run it from code and get the results in one call: POST https://api.apify.com/v2/acts/reportable_broth~weibo-scraper/run-sync-get-dataset-items?token=YOUR_TOKEN with the input above as the JSON body.

What does Weibo Scraper do?

• Keyword search, latest or top.
• A user's posts (newest first) and profile, fans and following.
• Comments on any post, sorted by likes or newest. Works deep without login.
• Weibo hot search (热搜) with heat values, and posts from the 热门 feed.
• Hot search changes: run it on a schedule and each topic is marked new, rising, falling, steady or dropped compared with your last run.
• A date range for keyword search and user posts, in Beijing time.
• Full text of long posts that lists cut off at "...全文".

Without login you get page 1 of search, timelines and follower lists. Page 1 is always the newest, so monitoring on a schedule works fine without a cookie.

What can I use it for?

• Watch what people say about a brand or product, every 15-30 minutes.
• Pull all the comments under one post to see how people reacted.
• Check what's on 热搜 right now, or track which topics climb and fall through the day.
• Pull everything posted about a keyword in one week (needs your cookie for more than the newest ~20).
• Get a creator's recent posts and follower count before you contact them.
• Collect Chinese-language posts on a topic for research.

How do I use it?

1. Open the actor on Apify and click Try for free. A free Apify account is enough to start.
2. Fill in what you want: keywords, users, or post URLs. You can mix them in one run.
3. If you need more than page 1 of search, user posts, fans or following, add your cookie (steps below). Skip this for comments, hot search and page-1 monitoring.
4. Click Start.
5. When it finishes, open the Output tab and download JSON, CSV or Excel. Or read the dataset through the API.

Getting your cookie

Only needed for page 2 and beyond.

1. In Chrome, open https://m.weibo.cn and log in (you can scan the QR code with the Weibo app).
2. Press F12, open the Network tab, press F5.
3. Click the first request (m.weibo.cn) and copy the whole Cookie value under Request Headers.
4. Paste it into cookie, with accountMode on own.

Apify stores it encrypted. It's only sent to Weibo.

Examples

Monitor keywords, no login (run every 15-30 min):

{ "searchKeywords": ["新能源汽车", "#比亚迪#"], "onlyNew": true, "maxCommentsPerFoundPost": 20 }

An account's posts and fans, with your cookie:

{ "users": ["https://weibo.com/u/2803301701"], "maxPostsPerUser": 200, "maxFans": 500, "cookie": "SUB=...; SUBP=...; ..." }

Comments on a post, plus what's trending:

{ "postUrls": ["https://weibo.com/2803301701/RjAABq8CG"], "maxCommentsPerPostUrl": 2000, "commentsSort": "latest", "hotSearch": true, "hotFeedPosts": 50 }

Track how the hot search board moves (run every 30 min):

{ "hotSearchChanges": true }

One week of posts on a keyword, with your cookie:

{ "searchKeywords": ["比亚迪"], "dateFrom": "2026-09-01", "dateTo": "2026-09-07", "maxPostsPerKeyword": 1000, "cookie": "SUB=...; SUBP=...; ..." }

Input

Field	Default
searchKeywords		Chinese, English or #hashtags#
maxPostsPerKeyword	50	~20 without login
searchSort	latest	latest (实时) or top (综合)
dateFrom, dateTo		2026-09-01 style, Beijing time. For keyword search and user posts
users	[]	User ID, profile URL or @name
includeProfile	true	Add a profile row per user
maxPostsPerUser	20	~10 without login
maxFans, maxFollowing	0	20 without login
postUrls	[]	Post URL or ID. Returns the post and its comments
maxCommentsPerPostUrl	100
maxCommentsPerFoundPost	0	Comments for each search/user post
commentsSort	hot	hot (most liked) or latest
hotSearch, hotFeedPosts	false, 0	Trending lists
hotSearchChanges	false	Compare the hot search board with your last run
fullText	true	Fetch full text of long posts
onlyNew	false	Skip items returned in earlier runs
accountMode	own	own uses your cookie (no cookie = no login), none never logs in
cookie		See above
requestGapSecs	0.8	Seconds between logged-in requests

What data do I get?

Every row has a type: post, comment, user, fan, following or hotSearch.

{
  "type": "post",
  "url": "https://weibo.com/2803301701/RjAABq8CG",
  "createdAt": "2026-09-18T07:38:24+00:00",
  "text": "...",
  "likes": 1520, "comments": 87, "reposts": 43,
  "region": "北京",
  "topics": ["比亚迪"],
  "user": { "userId": "2803301701", "name": "...", "verified": true, "followers": 120000 },
  "keyword": "新能源汽车"
}

Posts also have id, bid, source, mentions, images, video, isPinned, repostOf. Comments have floor and topReplies. Hot search rows have rank, word, heat, label. With hotSearchChanges on they also get status, previousRank, rankChange, peakRank, peakHeat, firstSeenAt and runsOnBoard. The first run has nothing to compare with, so every topic is baseline. Rounded counts like "1.2万" become numbers.

Text comes back as posted, mostly Simplified Chinese. Nothing is translated.

Limits

• Without login: page 1 only for search (~20), user posts (~10), fans and following (20). Comments and hot lists aren't limited.
• Hidden fan lists and deleted posts return nothing.
• The date range needs your cookie to reach older posts. Without one, only the newest ~20 are checked. Posts outside the range aren't saved and aren't charged. With searchSort: top the results aren't in time order, so the run can't stop early and may use more requests.
• Hot search changes are compared with the previous run of the same account, so only the gap between runs is covered. Topics that come and go between two runs are missed.
• Logged-in runs keep one IP. If Weibo rate-limits, it switches IP once and slows down. A second time and the run stops, so your account isn't pushed. If the login expires mid-run, it continues without login.

How much does it cost?

$4.49 per 1,000 items. Posts, comments, users, fans and hot topics each count as one. Rows from hotSearchChanges are $25 per 1,000, since each one carries the comparison with your earlier runs. There's no separate compute charge. With onlyNew on you only pay for new items.

Some real numbers:

• A hot search snapshot, 50 topics: 50 × $0.00449 = $0.22.
• Hot search changes every 30 minutes for a day: about 48 runs × ~55 rows = ~2,640 rows × $0.025 = about $66. Every hour halves that.
• 2,000 comments on one post, plus the post: 2,001 items = $8.98.
• 100 posts from a keyword with 20 comments each: 100 + 2,000 = 2,100 items = $9.43.

Apify's free plan gives you $5 of credit a month. That's about 1,100 items here.

FAQ

Why do I need a cookie? Weibo only shows page 2 and beyond of search, user posts, fans and following to logged-in users. Comments, hot search and the hot feed work without one.

Is my account safe? The run keeps one IP, waits 0.8 s between logged-in requests, and stops if Weibo rate-limits it twice. That keeps the load low, but I can't promise Weibo will never restrict an account. Use a secondary account if you can.

Why did I get so few results? Usually no cookie (page 1 only, about 20 posts), or onlyNew skipping what earlier runs already returned. Some users also hide their fan lists.

Can I schedule it? Yes. Save your input as a Task, add a Schedule, and turn on onlyNew so each run only returns new items.

Other scrapers I made

• RedNote / Xiaohongshu, Douyin, Bilibili
• Threads Scraper & Monitor, YouTube Monitor

Found a bug or need a field? Open an issue on the Issues tab.

中文说明

抓微博关键词搜索、用户微博、粉丝和关注、评论、热搜和热门微博。可以限定日期（北京时间），也可以追踪热搜变化（新上榜、上升、下降、掉榜）。不登录也能抓第一页、评论和热搜；要翻页就填自己的 cookie（步骤见上）。每 1,000 条 $4.49。

---

Only collects publicly available data. You're responsible for how you use it.