Pricing

from $0.85 / 1,000 rows

Try for free

Go to Apify Store

Weibo Scraper 微博 - Posts, Comments, Profiles, Search, No Login

Try for free

Returns Weibo (微博) posts with text, date, reposts, comments, likes, pictures and video links: an account's 100 newest posts without a cookie, its older posts with pictures or videos, any post by link with comments and replies, profiles, keyword search by date and the hot-search board. No login.

Pricing

from $0.85 / 1,000 rows

Rating

0.0

(0)

Developer

Ben

Actor stats

Bookmarked

Total users

Monthly active users

3 days ago

Last modified

What other Weibo scrapers state as their limits, and what this one does

The two most used Weibo Actors in the Store had no open or closed issue on October 6, 2026, so this table takes the limits from their own documentation.

Their documentation says	This Actor
An account's posts need your own login cookie, "Weibo serves timelines only to signed-in sessions"; the other Actor has no account posts at all	The 100 newest posts of an account without any cookie, and the 100 newest of each kind (original, video, picture, article). With `olderPosts` it continues with the older posts that have pictures or videos: 894 picture posts back to 2011 and 135 video posts of an account with 1,936 posts in the test
Accounts by numeric id only; "custom addresses such as weibo.com/name are not resolved"	Numeric id, profile link, custom address (`weibo.com/sinatech`) or the exact name (`新浪科技`)
No comments (one Actor); comments at the price of a post, $0.035 each (the other)	Comments and replies as rows of their own at $0.001: the 2,000 newest comments of one post in 83 seconds, replies with the comment they stand under
Search returns about 14 posts a page and at most 50 pages a run	50 posts per request, read back in time as far as the limit and the dates ask: 1,000 posts of one keyword in 88 seconds. A day five years back answers like a current one
Free plan: 25 rows a run, no runs through the API	The same Actor on every plan, through Console, API and schedules
Search rows carry no follower count of the author	`authorFollowers` is filled in search rows (rounded by Weibo above 10,000) and exact in an account's rows
One price for every row, comments and hot-search topics included: $0.035, or $0.019 to $0.075 depending on the plan	$0.004 a post or profile, $0.001 a comment, reply or hot-search topic

🔎 What is the Weibo Scraper?

Weibo is China's largest microblog: accounts of people, brands, media and authorities, a comment section under every post and a hot-search board that shows what the country talks about. Its open API is closed to developers outside China. This Actor makes the same data requests as Weibo's desktop and mobile sites, as a visitor who is not logged in, and returns clean rows.

What data does it extract?

Posts: whole text (long posts are completed), date in Beijing time, reposts, comments, likes, pictures in their largest size, video file link with cover and length, attached link, topics, posting device and region, pinned mark, and for a repost the quoted post with its own counts
Comments and replies: text, date, likes, number of replies, author name, id, followers and region, and for a reply the comment it stands under
Profiles: name, bio, exact followers, following and post counts, reposts, comments and likes received, verification with reason, location, creation date, birthday, company, credit level
Search results: the posts for a keyword, newest first back in time or Weibo's 50 hottest
Hot search: the board as it stands, with rank, heat value, label (热, 新, 沸, 爆), category and the time a topic entered the board

⬇️ Input

Fill in one or several of the first four fields.

Field	Type	Default	What it does
`accounts`	array		Numeric ids, profile links, custom addresses or exact names of accounts
`postUrls`	array		Post links, numeric ids or short codes: `https://weibo.com/1749127163/RlpYngQoW`, `https://m.weibo.cn/detail/5351047314014682`, `RlpYngQoW`
`searchQueries`	array		Keywords, Chinese or English
`hotSearch`	boolean	false	Adds the hot-search board
`maxResults`	integer	20	Posts per account and per search, topics of the board; up to 20,000
`accountOutput`	string	`posts`	`posts`, `profile` (one row per account, no posts) or `both`
`postType`	string	`all`	`all`, `original` (no reposts), `video`, `picture`, `article`: each is its own list of 100
`olderPosts`	boolean	false	Continue behind the newest 100 with the account's picture and video posts
`searchSort`	string	`newest`	`newest` (reads back in time) or `hot` (Weibo's 50 hottest posts for the words)
`postedAfter`, `postedBefore`	string		A day such as `2026-09-01` (Beijing time) or an age such as `7 days`; for searches and accounts
`fullText`	boolean	true	Complete posts that Weibo's lists cut at about 140 characters, without extra charge
`maxCommentsPerPost`	integer	0	Comments to save under each post, up to 20,000
`commentsSort`	string	`hot`	`hot` or `newest`
`includeReplies`	boolean	false	Replies under the comments, as rows of their own; they count towards the limit per post
`debug`	boolean	false	Log every request with its size and time

Example input

The newest posts of two accounts with their profiles:

{
  "accounts": ["新浪科技", "https://weibo.com/u/1749127163"],
  "accountOutput": "both",
  "maxResults": 100
}

Posts for a keyword from one week, newest first:

{
  "searchQueries": ["比亚迪"],
  "postedAfter": "2026-09-28",
  "postedBefore": "2026-10-04",
  "maxResults": 2000
}

Two posts with their 200 hottest comments and the replies:

{
  "postUrls": ["https://weibo.com/1749127163/RlpYngQoW", "https://m.weibo.cn/detail/5350984768294187"],
  "maxCommentsPerPost": 200,
  "includeReplies": true
}

The hot-search board:

{
  "hotSearch": true,
  "maxResults": 50
}

⬆️ Output

Every row has a type: post, comment, user or hot_search. The dataset has a table view for each. target says which input entry a row belongs to, and the key-value record SUMMARY lists per entry how many rows were saved and how it ended.

A post:

{
  "type": "post",
  "postId": "5351047314014682",
  "mid": "5351047314014682",
  "bid": "RlpYngQoW",
  "postUrl": "https://weibo.com/1749127163/RlpYngQoW",
  "text": "路上发现了几辆还没上牌的澎程N90 [酷]",
  "isTruncated": false,
  "createdAt": "Tue Oct 06 15:42:39 +0800 2026",
  "createdAtIso": "2026-10-06T15:42:39+08:00",
  "createdTimestamp": 1791272559,
  "source": "Xiaomi 18 Pro",
  "region": "北京",
  "repostsCount": 161,
  "commentsCount": 375,
  "attitudesCount": 3998,
  "commentsCurated": false,
  "authorId": "1749127163",
  "authorName": "雷军",
  "authorVerified": true,
  "authorFollowers": 27334341,
  "authorAvatar": "https://tvax4.sinaimg.cn/crop.0.0.1080.1080.1024/001Un9Srly8iguza3zz5rj60u00u0gp602.jpg?KID=…",
  "images": ["https://wx4.sinaimg.cn/large/001Un9Srly1ihq29slh3rj61400u07co02.jpg"],
  "videoUrl": null,
  "videoCover": null,
  "videoDurationSecs": null,
  "linkUrl": null,
  "linkTitle": null,
  "topics": [],
  "isPinned": false,
  "isRepost": false,
  "repostOf": null,
  "target": "account:雷军",
  "scrapedAt": "2026-10-06T22:11:39.368920+00:00"
}

A repost carries the quoted post in repostOf with its own postId, bid, postUrl, text, createdAt, createdAtIso, repostsCount, commentsCount, attitudesCount, authorId, authorName, images and videoUrl. The counts on the row itself are the repost's own.

A comment (a reply has isReply: true, the comment it stands under in rootCommentId, and replyToCommentId when it answers another reply):

{
  "type": "comment",
  "commentId": "5351048828944594",
  "text": "昨天我也遇到了，忍不住停下车拍[嘻嘻]我后天也要提探索版了，雷总[嘻嘻] http://t.cn/AXWezGcb",
  "createdAt": "Tue Oct 06 15:48:40 +0800 2026",
  "createdAtIso": "2026-10-06T15:48:40+08:00",
  "likeCount": 214,
  "replyCount": 13,
  "authorId": "6363461340",
  "authorName": "雪卡同学",
  "authorVerified": false,
  "authorFollowers": 2437,
  "authorRegion": "广东",
  "postId": "5351047314014682",
  "postUrl": "https://weibo.com/1749127163/RlpYngQoW",
  "isReply": false,
  "rootCommentId": null,
  "replyToCommentId": null,
  "target": "posts:post links",
  "scrapedAt": "2026-10-06T22:06:50.152305+00:00"
}

A profile:

{
  "type": "user",
  "userId": "2803301701",
  "screenName": "人民日报",
  "description": "人民日报法人微博。参与、沟通、记录时代。",
  "gender": "m",
  "location": "北京",
  "ipLocation": null,
  "followersCount": 158038626,
  "friendsCount": 3096,
  "statusesCount": 153968,
  "repostsReceived": 1184318157,
  "commentsReceived": 276638960,
  "likesReceived": 2528813908,
  "verified": true,
  "verifiedType": 3,
  "verifiedReason": "《人民日报》法人微博",
  "avatarHd": "https://tvax4.sinaimg.cn/crop.0.0.1018.1018.1024/0033ImPzly8h8vgemh8kxj60sa0sadgw02.jpg?KID=…",
  "coverImage": "https://wx1.sinaimg.cn/crop.0.0.640.640.640/0033ImPzly1gs64jwa7uyj60v90v9n1r02.jpg",
  "domain": "rmrb",
  "website": null,
  "profileUrl": "https://weibo.com/u/2803301701",
  "createdAt": "2012-07-22 02:28:35",
  "birthday": "1948-06-15 双子座",
  "company": "人民日报社",
  "sunshineCredit": "阳光信用极好",
  "target": "account:2803301701",
  "scrapedAt": "2026-10-06T22:08:29.445662+00:00"
}

A hot-search topic:

{
  "type": "hot_search",
  "rank": 1,
  "title": "兰香去世时没戴红绳",
  "hotValue": 303420,
  "labelName": "热",
  "category": "剧集",
  "isHot": true,
  "isNew": false,
  "isPromoted": false,
  "url": "https://s.weibo.com/weibo?q=%23%E5%85%B0%E9%A6%99%E5%8E%BB%E4%B8%96%E6%97%B6%E6%B2%A1%E6%88%B4%E7%BA%A2%E7%BB%B3%23",
  "onBoardSince": "2026-10-06T23:26:20+08:00",
  "target": "hot:hot search",
  "scrapedAt": "2026-10-06T22:07:51.280222+00:00"
}

💰 What a run costs

Apify plan	Per 1,000 posts or profiles	Per 1,000 comments, replies or hot-search topics
Free	$4.00	$1.00
Bronze	$3.80	$0.95
Silver	$3.60	$0.90
Gold, Platinum, Diamond	$3.40	$0.85

A post or profile row costs two events, the row ($0.001) and the record ($0.003); a comment, reply or hot-search row costs the row only. Apify's standard start event ($0.00005 per run at the default memory) is the only other charge. With a maximum charge set for the run, the Actor stops when the next row would pass it, and a post whose comments were started still gets its own row.

Examples on the Free plan: the 100 newest posts of an account, $0.40. 1,000 posts for a keyword, $4.00. One post with 2,000 comments, $2.00. The hot-search board, $0.05. 100 profiles, $0.40. A daily check of five accounts with 20 posts and 20 comments each (100 posts, about 1,600 comments), $2.00.

⏱️ Measured (October 6, 2026, on a test server outside Apify)

Run	Rows	Time
Sample input: the 20 newest posts of one account, long texts completed	20	12 s
Three accounts by name, link and custom address: profile and six original posts each	21	7 s
Ten profiles	10	12 s
The hot-search board *	50	5 s
20 posts by link *	20	8 s
Four post links with 25 comments and replies each	84	13 s
One post with its 2,000 newest comments *	2,001	83 s
One keyword, the 1,000 newest posts *	1,000	88 s
One keyword on one day of September *	300	71 s
150 posts for a keyword with up to 20 comments each	487	64 s
Five accounts, 20 posts each with 20 comments *	1,670	53 s
One account, 400 posts with `olderPosts` *	400	178 s

These times are from a server in Germany, not from Apify's network. Runs without a mark used the server's own address only; runs marked * also read through three Apify datacenter proxy addresses. Peak memory was 81 MB in the 487-row run. A run of 60 account posts, 60 search posts and one post link with 40 comments each was killed after 9 seconds with 169 rows stored and started again on the same dataset: it ended with 121 posts and 2,548 comments, none of them twice.

Comment depth: a post that Weibo counts 1,164 comments on gave 1,056 comments to the end of its list in 87 seconds (the count includes replies and comments Weibo hides from visitors). On a post with 12,343 comments a probe read 6,383 comments in 320 requests and was stopped there; deeper is not tested.

🤖 For AI agents

No credential is needed and every target field is an array. Smallest useful calls:

{"accounts": ["1642634100"], "maxResults": 20}

{"searchQueries": ["人工智能"], "maxResults": 50}

{"postUrls": ["https://weibo.com/1749127163/RlpYngQoW"], "maxCommentsPerPost": 50}

{"hotSearch": true}

Rows are told apart by type. A run with nothing to return ends as succeeded with zero rows and a status message that says why.

💡 Use cases

📊 Brand and topic monitoring in China: search a brand name every day and read the posts with their comments
🧑‍💼 KOL research: profiles with exact follower counts and the newest posts of a shortlist of accounts with their reposts, comments and likes
🔥 Trend tracking: the hot-search board on a schedule, to chart how a topic rises and how long it stays
🧠 Text data for analysis and model training: whole post texts and comment sections in Chinese, with author and region

⚠️ Limits, stated plainly

An account shows a visitor its 100 newest posts. Each kind (original, video, picture, article) is its own list of 100. Behind them, olderPosts returns the account's posts with pictures and its video posts as far back as the account shows them; text-only posts and reposts older than the newest 100 are not shown to a visitor and are not returned. The status message says so when an account has more.
Weibo's search is loose. In a one-day search for 小米汽车, 103 of 300 rows held the phrase, 68 both words apart, 111 one of them and 18 neither. For 比亚迪, 884 of 1,000 rows had the word in their text. Filter on text when you need exact matches.
A search reads back in time, 50 posts per request. A busy keyword has 1,000 posts in two days; a year of it is a long and costly run. Set postedAfter, postedBefore and maxResults to what you need. The hot order is 50 posts and no more.
Counts above one million: Weibo shows a visitor reposts and comments of a post as 1,000,000 at most (100万+). Likes are exact.
Comments: what a visitor can read. Weibo's comment count includes replies and hidden comments, so a complete list is shorter than the count. Where an author shows approved comments only (commentsCurated: true), a visitor gets those few or none.
authorFollowers is exact in an account's own rows, rounded by Weibo in search rows (to hundreds above 10,000) and empty for a post given by link. Ask for the profile to get the exact number.
Links that expire or need a header: video file links are signed and expire about an hour after the run; avatar links are signed as well and expire after a few hours. Picture links want the header Referer: https://weibo.com/ (HTTP 403 without it).
Hot search is the board at the moment of the run: 50 ranked topics plus the lines Weibo places on it (isPromoted: true, no rank). The pinned line above rank 1 is not included. There is no history; schedule runs to build one.
Speed: Weibo turns an address away for about three minutes after roughly 180 requests in a minute. The Actor sends two requests a second per address and reads through up to four addresses (its own and Apify datacenter proxies). A post with comments, a long post and an older picture post each cost a request of their own.
Not available without login: follower and following lists, an account's full timeline, Weibo's hot feed. ipLocation was empty for every account in the tests.
No memory between runs. A scheduled run returns the current newest posts again; drop the ones you have by postId.
Dates are Beijing time (UTC+8), in the input and in every row.

❓ FAQ

Do I need a Weibo account, a cookie or an API key? No. The Actor reads as a visitor who is not logged in.

How many posts of an account can I get? The newest 100, plus the newest 100 of each kind if you run the kinds one after the other, plus older posts with pictures or videos with olderPosts. Of an account with 1,936 posts the two older lists held 1,000 different posts in the test.

Can I search a period in the past? Yes. Set postedBefore to the last day of the period; the search starts there and reads back to postedAfter. A day in 2021 answered in the tests like a day of this week.

Why are there fewer comments than commentsCount? The count includes replies (switch on includeReplies) and comments that Weibo does not show to visitors. Authors can also limit a comment section to approved comments.

Are long posts complete? Yes, with fullText on (the default). Weibo's lists cut posts at about 140 characters; each cut post is completed with one more request and no extra charge. A row keeps isTruncated: true only when Weibo did not serve the rest.

I use another Weibo Actor. What changes? The input names mode, searchQuery, searchAliases, userIds, postIds, maxComments, includeComments, fetchFullText, searchDateFrom, searchDateTo, and operation, keyword, userId, weiboId, startDay, endDay, maxPages are read. Post, comment and profile rows use the field names postId, repostsCount, commentsCount, attitudesCount, authorName, likeCount, screenName, followersCount and so on. The differences: rows are told apart by type instead of mode; a value Weibo does not send is null; the names weiboId, repostCount, postedAt and picUrls are not written. Not offered: follower and following lists, the hot feed, Black Cat complaints, sentiment scores, province rollups, automatic brand translation and only-new runs. A run that asks for one of the first three says so and saves nothing.

What happens when an account or post does not exist? The run succeeds, the status message and SUMMARY name it, and nothing is charged for it.

Can I get the pictures and videos themselves? The rows hold the links. Fetch pictures with the header Referer: https://weibo.com/, and videos within the hour.

Is it legal? The Actor reads only what Weibo shows to any visitor without an account. Posts and comments contain personal data of their authors: use them for a lawful purpose, respect GDPR, China's PIPL and the rules of your own country, and do not use the data for spam or to target individuals.

🔗 You might also like

Bilibili Scraper: videos, comments, creators and rankings of Bilibili
Threads Public Scraper: posts of public Threads accounts
Telegram Channel Scraper: posts of public Telegram channels
TikTok Comments Scraper: comments and replies of TikTok videos

Keywords: weibo scraper, weibo api, sina weibo scraper, 微博爬虫, 微博数据, weibo posts scraper, weibo comments scraper, weibo user profile, weibo search api, 微博热搜, weibo hot search, weibo trending, china social media scraper, chinese social listening, weibo kol monitoring, weibo no login, weibo no cookie, weibo data export

Weibo User Scraper 微博 | Timelines, Profiles & Followers

zen-studio/weibo-user-scraper

Collect any Weibo (微博) account's posts with engagement counts, images, video and the province each was posted from. Add the full account profile, and the accounts it follows or is followed by. Paste a profile link, a username or an account ID.

Zen Studio

Weibo(微博) Media Downloader 🎥 📸

easyapi/weibo-wei-bo-media-downloader

Extract and download high-quality videos and images from Weibo (微博) posts. Supports multiple quality options up to 2K resolution, batch processing, and provides structured output with direct download links. Perfect for content archiving and social media research.

EasyApi

187

Weibo Scraper - User / Feed / Post

saswave/weibo-scraper-user-feed-post

$0.8 / 1000 results. Extract data from feed, post or user profile. Followers, likes, comments, title, content, media and more

SASWAVE

1.0

(1)

Weibo Scraper 微博 | Keyword Search, Date Range & Filters

zen-studio/weibo-search-scraper

Scrape Weibo (微博) posts by keyword with date range and filters: likes, comments, reposts, location, hashtags, images, video and author profiles. No login.

Zen Studio

Weibo Scraper API - Search, Comments & Brand Monitoring

webdata_labs/weibo-scraper

Search public Weibo posts and extract comments, authors, engagement, evidence URLs, and optional English translations. No login fields.

WebData Labs

Weibo Hot Search Tracker · 微博热搜 API & Trend Monitor

nexgendata/weibo-hot-search-tracker

Track the Weibo Hot Search board (微博热搜) live — China's #1 social-trending barometer. Top-50 ranked topics with heat score, category, 新/热/沸 flags, pinyin & best-effort English translation. A Hootsuite-style China social-listening signal for brand, PR & consumer-research teams. No login.

NexGenData

Weibo Comments Scraper 微博 | Replies, Likes & Profiles

zen-studio/weibo-comments-scraper

Collect every comment on a Weibo (微博) post with likes, reply counts, the province each commenter wrote from, images, and their profile. Optionally pull threaded replies as their own rows. Paste any post link.

Zen Studio

Truth Social Scraper - Profiles & Posts

benthepythondev/truth-social-scraper

Scrape public Truth Social posts and profile metadata by username. Export post text, engagement, media, links, tags, timestamps and account statistics.

Ben

Weibo Scraper | 微博 Profiles & Posts: Reposts, Comments, Likes

themineworks/weibo-scraper

Scrape Weibo (China's dominant Twitter-equivalent) by user ID: follower count, verification, bio, and full post timeline with reposts, comments and likes. China residential proxy, no login, no browser.

The Mine Works

Weibo Scraper

lergassy/weibo-scraper

[from $3/1K posts] Scrape Weibo posts from accounts, keyword search and post links: full text, likes, comments, reposts, pictures, video, IP region, author. Add comments, profiles and the Weibo hot search list. No login, no cookies, no run fee.