Weibo 微博 Scraper - Hot Search, Posts, Comments & Users

Weibo Scraper extracts public data from Weibo 微博 (weibo.com, also known as Sina Weibo) and returns clean, flat JSON. It covers the live hot search list (微博热搜), keyword search, advanced search by date range, post details, public comments and every reply to a comment, user profiles with recent posts, public following lists, reposts, post reactions and user search. Public profile fields of the users in the results are included, such as user ID, avatar, IP location (IP属地), gender and verification. No Chinese-language skills, proxies, cookies or login are needed, and you can use it as a Weibo API: call it over HTTP or with the Apify Python or JavaScript client and get the same JSON back.

What is Weibo?

Weibo (微博), often called Sina Weibo, is a large Chinese public social network built around short posts. People, brands, media and public figures publish posts with text, images and video, and other users comment, repost and react. Its hot search list (热搜) shows which topics are trending at the moment.

What data can you get?

The Actor has 17 modes. Each mode takes a different input and returns a different item type.

Mode	Give it	You get
hot	nothing	the current trending list: rank, keyword, heat (hotness), label, link
search	a keyword	posts: text, author with ID, avatar, verification and IP location, likes, comments, reposts, publish time, link
advancedSearch	a keyword, optional startDate and endDate	posts in that date range: text, author, source, counts, image and video flags, link
post	post links or IDs	the same post fields for each post
comments	post links or IDs	comment text, likes, reply count, time, plus each commenter's public profile: user ID, nickname, avatar, IP location, gender, location, bio, followers, verification. includeReplies adds the first replies under each comment (off by default, billed like comments, long threads may be partial)
replies	comment IDs (commentId from a comments run)	every reply under a comment, with each replier's public profile and IP location
reposts	post links or IDs	each repost: text, reposter's public profile, IP location, time, counts, link
likes	post links or IDs	the people who reacted to a post (about 50 per post, the platform's own limit), with reaction type and public profile
user	UIDs or profile links	profile (nickname, avatar, gender, location, bio, verification, followers, post count) plus recent posts
following	UIDs or profile links	the accounts it follows, with profile fields, newest follows first
userSearch	a keyword, optional province	matching accounts: user ID, nickname, avatar, follower count, link. province limits it to one province
topicSearch	a keyword	matching topics (hashtags): name, description, discussion and read counts, cover, link
videoSearch	a keyword	video posts: text, author, source, counts and the direct video link
picSearch	a keyword	pictures: full-size and small image links, the post text and ID, author, time
userVideos	UIDs or profile links	the user's videos: text, play count, likes, length in seconds, cover and direct video link
userPhotos	UIDs or profile links	the photos of the user's album: image links with the post each belongs to
userArticles	UIDs or profile links	the posts in the user's article tab (long posts); headline articles carry articleTitle and articleUrl

Every item has a type field: post, comment, user, hot, following, repost, liker, userResult, topic, picture, video, photo or article.

The user mode also takes onlyOriginal: only the posts the user wrote themselves (no reposts), about 49 per page. A post that has a video now also carries its play count, length and video link.

Post and repost fields (search, advancedSearch, post, user, reposts modes)

Field	Description
text	Post text in the original language
likes, comments, reposts	Engagement counts
publishedAt	Publish time
authorName, authorId, authorUrl	Author nickname, numeric user ID and profile link
authorIpLocation	IP location shown with the post (published in), e.g. a province
authorAvatar	Author's avatar image URL (largest available)
authorVerified, authorVerifiedType	Whether the author is verified, and the platform's verification type code (-1 or 220 for none)
postId, mblogid, url	Numeric post ID, short post ID and post link
imageCount	Number of images
isPinned	Whether the post is pinned
originalPostId	reposts only (repost items): the post that was reposted
publishedAtText	advancedSearch only: publish time as Weibo's search page shows it (for example 10月09日 23:59, Beijing time, no year)
source	advancedSearch only: where the post was published from, such as a super topic or app name
postKind	advancedSearch only: original, repost and so on, as the search page labels it
hasVideo	Whether the post has a video (advancedSearch, videoSearch and the post-based modes)
videoPlays, videoDurationSeconds, videoUrl	Video play count, length and direct video link, for posts with a video (post, user and article items; empty without a video). Video links expire after some hours, so download soon

Topic, picture, video, photo and article fields (topicSearch, picSearch, userVideos, userPhotos, userArticles modes)

Field	Item types	Description
keyword, url, description	topic	Topic name, link to its search page and description
discussions, reads	topic	Discussion and read counts of the topic
coverUrl	topic, video	Cover image link
imageUrl, thumbnailUrl	picture, photo	Full-size and small image links
postId, url, text, publishedAt	picture, photo, video, article	The post the picture, photo or video belongs to (a picture's link may be empty)
authorName, authorId, authorUrl, authorAvatar	picture	Who posted it (video and photo items carry the author ID and, for videos, the name)
videoPlays, videoDurationSeconds, videoUrl, likes	video	Play count, length in seconds, direct video link and like count
publishedAtText	video	Publish date as the profile page shows it (month-day, no year)
articleTitle, articleUrl	article	Headline article title and link; empty for ordinary long posts in the tab. article items carry all the post fields

Comment and reply fields (comments, replies modes)

Field	Description
text	Comment text
likes, replyCount	Likes and number of replies
publishedAt	Comment time
authorName, authorId	Commenter nickname and numeric user ID
authorIpLocation	IP location shown with the comment (from), e.g. a province
authorGender	Commenter's gender: male or female
authorLocation	Location the commenter shows on their profile
authorBio	Commenter's bio as published
authorFollowers	Commenter's follower count
authorAvatar	Commenter's avatar image URL
authorVerified, authorVerifiedType	Verification flag and type code
commentId	Comment ID (use it as input for replies mode)
parentCommentId	ID of the top-level comment a reply belongs to; null on top-level comments
postId	The post the comment is under

User profile fields (user mode)

Field	Description
uid, nickname, url	User ID, nickname and profile link
followers, following, postCount	Follower, following and post counts
verified, verifiedType, verifiedReason	Verification flag, type code and verification text
bio	Bio as published
gender	male or female
location	Location shown on the profile
avatarUrl	Avatar image URL (largest available)

Trending fields (hot mode)

Field	Description
rank	Position on the hot search list (empty for sponsored slots)
keyword	Trending topic
hotness	Heat value
label	Label shown next to the topic
url	Link to the topic search

Following, reactions and user search

Field	Item types	Description
uid, nickname, avatarUrl, followers, url	following, liker, userResult	Public profile basics
bio, gender, location, verified, verifiedType, verifiedReason, following, postCount	following, liker	The same profile fields as user mode; some are often empty (see Tips and limits)
ofUserId	following	The account whose following list this is
reaction	liker	Reaction type: like, haha, surprised, sad, angry, tip or hug
likedPostId	liker	The post that was reacted to, as you gave it

The full field list is in the dataset schema on the Actor page.

How to use Weibo Scraper

1. Open the Actor on Apify and click Try for free (or Start).
2. Choose a mode, for example Trending list (hot search) or Post comments.
3. Fill in the input for that mode: a keyword, post links, user IDs or comment IDs. The hot mode needs no input.
4. Set maxResults to the number of items you want in total.
5. Click Start and wait for the run to finish.
6. Open the dataset and export it as JSON, CSV or Excel, or fetch it through the API.

Input

Field	Type	Required	What it does	Default
mode	string	required	What to scrape: search, post, comments, user, hot, following, reposts, likes, replies, userSearch, advancedSearch, topicSearch, videoSearch, picSearch, userVideos, userPhotos, userArticles	search
keyword	string	optional	Search keyword for search, userSearch, advancedSearch, topicSearch, videoSearch and picSearch. Chinese keywords return the best results.	原神
posts	array of strings	optional	Weibo post links or IDs for post, comments, reposts and likes, e.g. https://weibo.com/6593199887/Rlf5vlWRz or 5350628895231161	none
users	array of strings	optional	Weibo UIDs or profile links for user, following, userVideos, userPhotos and userArticles, e.g. 6593199887 or https://weibo.com/u/6593199887. Accounts that hide their following list return nothing and are not billed.	none
includeReplies	boolean	optional	Comments mode: also return the replies shown under each comment. Each reply is billed like a comment.	false
onlyOriginal	boolean	optional	User mode: list only the user's own posts (no reposts).	false
province	string	optional	userSearch: only accounts located in this province, e.g. 浙江. Empty means all regions.	empty
commentIds	array of strings	optional	Replies mode: numeric comment IDs, e.g. the commentId of a comment from comments mode. Returns every reply under each comment.	none
startDate	string	optional	advancedSearch: first day to search, like 2026-10-01. Set together with the end date, or leave both empty to search all time.	none
endDate	string	optional	advancedSearch: last day to search (the whole day is included), like 2026-10-10	none
maxResults	integer	optional	Maximum number of items to return in total for this run (1 to 5,000). You are charged per item returned.	20

maxResults is the total number of items for the whole run. The run stops as soon as it is reached, or when the spending limit you set in the run options is used up.

Trending list:

{ "mode": "hot", "maxResults": 50 }

Keyword search:

{ "mode": "search", "keyword": "原神", "maxResults": 50 }

Comments with replies:

{ "mode": "comments", "posts": ["https://weibo.com/6593199887/Rlf5vlWRz"], "includeReplies": true, "maxResults": 200 }

Following list:

{ "mode": "following", "users": ["6593199887"], "maxResults": 500 }

Advanced search by date range:

{ "mode": "advancedSearch", "keyword": "原神", "startDate": "2026-10-01", "endDate": "2026-10-10" }

With "includeReplies": true in comments mode, the replies the platform shows under each comment are returned right after their comment, with parentCommentId set. Replies are billed like comments and count toward maxResults; long reply threads may be partial. To get every reply under a comment, use replies mode with the commentId.

Output example

Post item (values are illustrative):

{
  "type": "post",
  "postId": "5350628895231161",
  "mblogid": "Rlf5vlWRz",
  "url": "https://weibo.com/6593199887/Rlf5vlWRz",
  "text": "Example post text",
  "authorName": "ExampleAccount",
  "authorId": 6593199887,
  "authorUrl": "https://weibo.com/u/6593199887",
  "authorAvatar": "https://example.com/avatar.jpg",
  "authorVerified": true,
  "authorVerifiedType": 0,
  "authorIpLocation": "上海",
  "publishedAt": "2026-10-05T04:00:01Z",
  "reposts": 51,
  "comments": 26,
  "likes": 612,
  "imageCount": 0,
  "isPinned": false
}

Comment item:

{
  "type": "comment",
  "postId": "5350628895231161",
  "commentId": "5350631234567890",
  "parentCommentId": null,
  "text": "Example comment text",
  "authorName": "ExampleUser",
  "authorId": 1234567890,
  "authorAvatar": "https://example.com/avatar2.jpg",
  "authorVerified": false,
  "authorVerifiedType": -1,
  "authorIpLocation": "广东",
  "authorGender": "female",
  "authorLocation": "广东 广州",
  "authorBio": "Example bio",
  "authorFollowers": 328,
  "publishedAt": "2026-10-05T05:12:40Z",
  "likes": 14,
  "replyCount": 2
}

User profile item:

{
  "type": "user",
  "uid": 1234567890,
  "nickname": "ExampleAccount",
  "url": "https://weibo.com/u/1234567890",
  "avatarUrl": "https://example.com/avatar3.jpg",
  "gender": "male",
  "location": "北京",
  "bio": "Example bio",
  "verified": true,
  "verifiedType": 0,
  "verifiedReason": "Example verification text",
  "followers": 125000,
  "following": 310,
  "postCount": 4210
}

Trending item:

{
  "type": "hot",
  "rank": 1,
  "keyword": "示例话题",
  "hotness": 1250000,
  "label": "热",
  "url": "https://s.weibo.com/weibo?q=example"
}

How much does it cost?

The Actor uses pay-per-event pricing: a small start fee per run plus a fee per item returned.

Event	What it covers	Paid plans
Run start	charged once per run	$0.005
Result	one post, comment, reply, repost, reaction, following entry, search hit (post, user, topic, picture or video), trending entry, video, photo or article	$0.004
Profile	one user profile	$0.0092 (Bronze), $0.0083 (Silver), $0.0075 (Gold, Platinum, Diamond)

Example: a search run on a paid plan that returns 100 posts costs $0.005 + 100 x $0.004 = about $0.41. A userVideos run for one account that returns 45 videos costs $0.005 + 45 x $0.004 = about $0.19.

On the free Apify plan, every run returns a sample dataset: a real result of this Actor's prefilled input, captured on the date shown in the run status, so you can see exactly what the output looks like. Your own input is not used and nothing is charged. To run your own searches, use any paid Apify plan.

The start fee is charged once per run, when the first request to the data source succeeds, even if that request returns nothing. It is not charged when the data source fails from the start or rejects all the post, user or comment IDs you gave, or when the spending limit cannot cover it. Items are charged only when they are actually returned. Set a maximum cost per run in the run options and the Actor stops when it is reached.

Use cases

• Brand and PR monitoring: track mentions of your brand, product or competitors and read what people say in the comments and replies.
• Trend spotting: pull the hot search list (微博热搜) on a schedule and see which topics rise and fall, with rank and heat.
• Influencer research: compare followers, verification and post engagement of accounts in your niche, and find accounts with user search.
• Audience reaction analysis: see who reposted or reacted to a post, which reaction types they used, and the IP locations (provinces) commenters post from.
• Event and campaign research: use advanced search with a date range to collect posts about a launch or event in a specific period.
• AI and NLP datasets: collect Chinese social posts and comments for sentiment analysis or retrieval.

Integrations and API

You can start runs through the Apify API, run the Actor on a schedule, trigger webhooks when a run finishes, and connect it to integrations such as Zapier, Make and Google Sheets. It also works with AI agents through the Apify MCP server. Python example:

from apify_client import ApifyClient
client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("clover_folktale/weibo-scraper").call(run_input={"mode": "hot", "maxResults": 50})
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item)

Another Python example for the trending list: weibo_trending.py.

Tips and limits

• Pagination: raise maxResults (up to 5,000 per run); pagination is handled for you.
• Free plan. Every run on the free Apify plan returns a stored sample dataset (a real result of the prefilled input); your own input is not used and nothing is charged. Use any paid Apify plan to run your own searches.
• User posts: authorIpLocation is empty for posts from user mode because Weibo's profile timeline doesn't carry it. Search results and post mode do include it.
• Following lists: many accounts hide their following list, and Weibo caps how deep a list goes (roughly the first 1,000 entries). A hidden or empty list returns nothing and is not billed; the run reports it as nothing found rather than failing. In the following list, bio, location and verifiedReason are empty for accounts that never filled them in, and postCount is empty.
• New list modes: userVideos, userPhotos and userArticles return what Weibo's profile tabs show, newest first, and stop when the tab ends. userArticles lists the posts of the article tab: only some of them are headline articles with a title (articleTitle); the rest are long posts. userPhotos returns one item per photo, so one post with several photos gives several items. picSearch links can expire like Weibo's own image links, so download what you need.
• Reposts: text is often just the platform's default "转发微博" when the reposter added no comment.
• Reactions (likes mode): the platform returns one page of about 50 per post, with no further pages.
• Advanced search: publishedAtText is the date as Weibo's search page prints it (no year, no seconds); there is no ISO timestamp, IP location or verification flag. startDate and endDate cover whole days.
• Replies: postId is empty when the platform doesn't say which post the comment is under. With includeReplies in comments mode, long reply threads may be partial; use replies mode for every reply.
• Hot search: rank is empty for sponsored slots, which Weibo doesn't number.
• Search volume: search returns the posts Weibo ranks for the keyword; the number per page varies, so you may see fewer posts than on weibo.com.
• Language: post text and comments are returned in the original language. Chinese keywords give the best search results.

FAQ

Do I need a Weibo account or cookies? No. You don't need a login, cookies or proxies.

Is it legal? Does it collect personal data? The Actor only returns publicly visible information. That includes the public profile of users who appear in the results, such as user ID, avatar, IP location, gender, location, verification and bio, as published on the platform. Private messages and anything not publicly visible are never collected. Bios are returned as published, including any business contact info the creator put there. You are responsible for using the data in line with applicable laws and Weibo's terms.

How fresh is the data? Data is collected when you run the Actor. The hot mode returns the trending list as it is at the time of the run.

Why are some fields empty? Some fields are only available in certain modes, or are left empty by the platform or the user. See Tips and limits above for the full list.

How many results can I get? Up to 5,000 items per run with maxResults. Some lists have platform limits: about 50 reactions per post and roughly 1,000 entries for a following list.

Can I schedule runs and export the data? Yes. Like any Apify Actor, it can run on a schedule, be started via the API, and send results to integrations. You can export each dataset as JSON, CSV or Excel.

Are results in Chinese? Post text and comments are returned in the original language. Field names are in English.

Something looks wrong or a field is empty. What should I do? Open an issue on the Actor page with the run link; the Actor is monitored daily.

More tools from the same developer

• Douyin 抖音 Scraper - Videos, Comments, User IDs & IP Location
• Xiaohongshu 小红书 RedNote Scraper - Notes, Comments & User IP
• Zhihu 知乎 Scraper - Answers, Articles, Comments & Users
• Bilibili B站 Scraper - Videos, Comments, Subtitles & User IDs
• WeChat Official Account 公众号 Scraper - Articles & Read Counts
• Douban 豆瓣 Scraper - Movies, Ratings, Comments & Reviews