Weibo Scraper - Search, Accounts, Comments & Hot List

Search Weibo, read an account's latest posts, pull the comments under a post, or take the real-time hot search list: one row per post, comment or hot topic, with the text, the author and their follower count, likes, comments, reposts, pictures, video links and the place Weibo says it was posted from.

Weibo shows people who aren't signed in one page of results per search tab and an account's latest few dozen posts, and this actor doesn't sign in. So a search brings back 50 to 90 posts and an account about 30, however many exist. Comments don't have that limit: one test run took 2,000 from a single post in under two minutes.

Input	Search terms, accounts (id, profile link or screen name), post links, or the hot search list switch
Output	One row per post, comment or hot topic: text, author, follower count, likes, comments, reposts, pictures, video link, place, time
Ceiling	About 50 to 90 posts per search and about 30 per account, which is what Weibo shows without signing in; 2,000 comments per post; 100 entries in each list per run
Speed	10 to 20 seconds a search, 5 to 11 seconds an account, about 1,100 comments a minute
Account needed	None from you
Price	$1.85 per 1,000 posts, $4.60 per 1,000 comments, $4.60 per 1,000 hot topics, flat on every plan. The free plan's $5 a month covers about 2,700 posts

🔍 What Weibo Scraper does

Four jobs, and one run can do all of them. Each runs only when you fill in its field.

• Search. Each term is looked up across Weibo's five result tabs (top, latest, hot, video, pictures), and a post found on two tabs is kept once.
• Accounts. An account's latest posts, newest first. Only its own, reposts it made included.
• Comments. The comments under a post, newest first or in Weibo's hot order, with their replies if you want them.
• Hot search list. The ranked list as it stands when the run reads it: about 50 topics with Weibo's heat figure.

Weibo cuts long posts short in its lists, so the whole text is fetched for each one. Every row is checked against what you asked before it goes in. A search post has to be a result of that search, an account's post has to be that account's, a comment has to belong to that post. Anything else is left out and not charged.

📋 What data you get from each Weibo post

What you get	Field
The post's id, short id and link	id, bid, url
When it was posted, in UTC	createdAt
The whole text, and whether it is whole	text, textIsComplete
The author, their verified label and follower count	authorId, authorName, authorUrl, authorVerified, authorVerifiedReason, authorFollowers, authorFollowersText
Likes, comments and reposts	likes, comments, reposts, countsRounded
Topics, @mentions and outside links in the text	topics, mentions, links
Pictures, and how many the post has	images, imageCount
Video link, cover, length and plays	videoUrl, videoCoverUrl, videoDurationSeconds, videoPlays
The post it reposts	isRepost, repostOf
The place Weibo prints on it, and the line beside its time	postedFrom, sourceLabel
Pinned, and how many times it was edited	isPinned, editCount
Which search or account it came from	source, searchTerm, searchTab, accountInput
A comment: its post, text, likes, replies, author and place	postId, postUrl, text, likes, replies, parentCommentId, replyToName, authorName, authorFollowers, authorLocation, postedFrom
A hot topic: rank, words, heat, category, label, time on the list	rank, topic, heat, category, label, isPinned, enteredListAt, searchUrl

▶️ How to scrape Weibo

1. Open Weibo Scraper and click Try for free.
2. Type words into Search terms, accounts into Accounts, post links into Posts for comments, or tick Hot search list. Any mix works.
3. Lower Posts per search, Posts per account or Comments per post while you're testing.
4. Click Start.
5. Download the rows as JSON, CSV or Excel, or read them through the Apify API.

Start it with every field empty to see the row shape first: one labelled sample post, not charged.

💰 How much does it cost to scrape Weibo?

$1.85 per 1,000 posts, $4.60 per 1,000 comments and $4.60 per 1,000 hot topics. Flat on every Apify plan, no volume tiers. A reply counts as a comment. On the free plan, the $5 Apify gives you each month covers about 2,700 posts, or about 1,050 comments.

You pay for rows that come back. The sample row and every note row are free, and so are posts the checks left out, repeats, and the promoted slot in the hot list. If you set a maximum charge for the run, it stops before the first row it couldn't pay for.

📥 What you give it

{
  "searchTerms": ["咖啡", "华为 手机"],
  "users": ["1749127163", "https://weibo.com/rmrb"],
  "postUrls": ["https://weibo.com/2803301701/RkyFjzhSX"],
  "maxCommentsPerPost": 200,
  "hotList": true
}

Field	Default	What it is
searchTerms	none, the box starts at 咖啡	Words to search for, one per line, in any language. Up to 100.
maxPostsPerSearch	50	The most posts from one search, 1 to 100. Weibo rarely has more than 90 to show.
matchAllWords	true	Keeps only posts whose text, reposted post or topic card holds every word you searched, split on spaces. Off returns everything Weibo's search does.
users	none	Account ids, profile links (weibo.com/u/..., weibo.com/n/<name>, weibo.com/<custom address>, m.weibo.cn/u/...) or screen names. Up to 100.
maxPostsPerUser	30	The most posts from one account, newest first, 1 to 50.
postUrls	none	Post links (weibo.com/<account id>/<post id>, m.weibo.cn/detail/<id>) or post ids. Up to 100.
maxCommentsPerPost	100	The most comments from one post, 1 to 2,000. Replies come on top.
commentsOrder	newest	newest goes through every comment Weibo lists. hot follows Weibo's ranking, which ends near 300.
includeReplies	false	Also returns the replies under each comment.
maxRepliesPerComment	20	The most replies under one comment, 1 to 100.
hotList	false	Returns the real-time hot search list.
fullText	true	Fetches the whole text of posts Weibo cuts short. Off is quicker and leaves textIsComplete false on those rows.

Two entries naming one account, an id and a screen name say, are read once. A setting outside its range stops the run before anything is fetched. An entry that isn't a Weibo link or id is skipped with a note, and the rest go ahead.

📤 What you get back

A real post from a search for 华为 手机 on 3 October 2026:

{
  "recordType": "post",
  "id": "5350042053381127",
  "bid": "RkZOZebAj",
  "url": "https://weibo.com/1278131290/RkZOZebAj",
  "createdAt": "2026-10-03T13:08:07.000Z",
  "text": "刚刷到一家华为门店在Mate90首销日卖了231台手机[哆啦A梦吃惊][哆啦A梦吃惊]",
  "textIsComplete": true,
  "authorId": "1278131290",
  "authorName": "搞机大师兄",
  "authorUrl": "https://weibo.com/u/1278131290",
  "authorVerified": true,
  "authorVerifiedReason": "数码博主",
  "authorFollowers": 448000,
  "authorFollowersText": "44.8万",
  "reposts": 0,
  "comments": 27,
  "likes": 55,
  "countsRounded": false,
  "topics": [],
  "mentions": [],
  "links": [],
  "images": ["https://wx3.sinaimg.cn/mw2000/4c2ebc5aly1ihpgye4uzlj210o274h0k.jpg"],
  "imageCount": 1,
  "videoUrl": null,
  "videoCoverUrl": null,
  "videoDurationSeconds": null,
  "videoPlays": null,
  "isRepost": false,
  "repostOf": null,
  "isPinned": false,
  "postedFrom": "山东",
  "sourceLabel": "数码博主",
  "editCount": 0,
  "source": "search",
  "searchTerm": "华为 手机",
  "searchTab": "top",
  "accountInput": null,
  "scrapedAt": "2026-10-03T15:08:27.798Z"
}

A comment and a hot topic from the same afternoon:

{
  "recordType": "comment",
  "id": "5350030405013535",
  "postId": "5348998178410235",
  "postUrl": "https://weibo.com/2803301701/RkyFjzhSX",
  "parentCommentId": null,
  "replyToName": null,
  "createdAt": "2026-10-03T12:21:50.000Z",
  "text": "祝祖国繁荣昌盛，国泰民安！",
  "likes": 1,
  "replies": 0,
  "authorId": "7911007775",
  "authorName": "无尽夏-Cecily",
  "authorUrl": "https://weibo.com/u/7911007775",
  "authorVerified": true,
  "authorFollowers": 2753,
  "authorLocation": "北京",
  "postedFrom": "江苏",
  "imageUrl": null,
  "commentsOrder": "newest",
  "postInput": "https://weibo.com/2803301701/RkyFjzhSX",
  "scrapedAt": "2026-10-03T15:11:40.918Z"
}

{
  "recordType": "hotTopic",
  "rank": 1,
  "topic": "爬珠峰的人都堵了",
  "heat": 1686010,
  "category": "幽默",
  "label": "热",
  "isPinned": false,
  "enteredListAt": "2026-10-03T12:48:21.000Z",
  "searchUrl": "https://s.weibo.com/weibo?q=%23%E7%88%AC%E7%8F%A0%E5%B3%B0%E7%9A%84%E4%BA%BA%E9%83%BD%E5%A0%B5%E4%BA%86%23",
  "postUrl": null,
  "listReadAt": "2026-10-03T15:07:58.138Z",
  "scrapedAt": "2026-10-03T15:07:58.138Z"
}

Field	How to read it
authorFollowers	Weibo rounds big accounts in its lists, so 448,000 here was "44.8万". authorFollowersText is the figure as Weibo printed it.
countsRounded	true when Weibo gave a post's likes, comments or reposts rounded, such as "100万+".
postedFrom	The province or country Weibo prints on a post or comment for where it was sent from. null when it prints none.
authorLocation	The place a commenter wrote on their own profile, which can differ from postedFrom.
sourceLabel	The line Weibo prints beside the time: the app or phone on an account's posts, the author's label in search results.
imageCount	How many pictures the post has. images holds the ones Weibo lists, nine at most.
videoUrl	Stops working an hour after the row was read, so download a video you want to keep. Picture links don't expire that way.
searchTab	The result tab a post was first found on: top, latest, hot, video or pictures.
rank, heat	A topic's place on the list and Weibo's heat figure. The pinned topic above the list has neither.

🧾 Reading the output

Row	How to spot it	Charged
A post	recordType: "post"	yes, as a post
A comment or a reply	recordType: "comment"; a reply has parentCommentId	yes, as a comment
A hot topic	recordType: "hotTopic"	yes, as a hot topic
The sample row	recordType: "sample", _sample: true	no
A note	recordType: "diagnostic", _diagnostic: true, an errorCode	no

errorCode	What happened
NO_RESULTS	Weibo has no posts for that search.
NO_MATCHES	Weibo found posts, but none hold every word of the search.
WITHHELD	Weibo doesn't show results for that search.
USER_NOT_FOUND	No account matches that id, link or name.
USER_UNAVAILABLE	Weibo says the account can't be viewed, usually because it was suspended.
NO_POSTS	The account shows no posts to people who aren't signed in.
POST_NOT_FOUND	The post doesn't exist or was deleted.
POST_HIDDEN	Only some readers can see the post, so its comments can't be read.
NO_COMMENTS	The post has no comments, or Weibo lists none.
PARTIAL	Weibo stopped answering partway. What came before is in the dataset.
NOT_ANSWERED	Weibo didn't answer for that entry this time. Run it again later.
NOT_REACHED	The run stopped first, at your maximum charge or its time limit.
BAD_INPUT	An entry or a setting that can't be used.
MAX_CHARGE_TOO_LOW	Your maximum charge doesn't cover one row, so nothing was fetched.
SETUP	Skipped because of a problem on our side.

To keep only the data, filter on recordType being post, comment or hotTopic. RUN_REPORT in the run's key-value store lists every entry with its status and how many rows it gave.

💡 What people use it for

• Hearing what Chinese buyers say about a brand. Schedule a daily search for its name and keep the post ids you haven't seen.
• Following newspapers, officials or creators without the app: their latest posts with likes and comments, newest first.
• Reading the reaction under one post, a launch or a public apology, 2,000 comments deep.
• Watching what trends. Read the hot list every hour and you have each topic's rank and heat across the day.

From a trending topic to what people say about it, in three runs:

1. Tick Hot search list and pick topics from the topic column.
2. Put them into Search terms and sort the posts on comments.
3. Put the url of the posts you want into Posts for comments.

🚧 What it does not do

• No full history. An account gives its latest 30 or so posts and a search 50 to 90, because that's what Weibo shows without signing in. It doesn't sign in.
• No date range on searches. Weibo's own advanced search asks you to sign in. Filter on createdAt afterwards.
• No follower lists, no lists of who liked or reposted.
• Nothing hidden from the public. Posts limited to friends or fans stay hidden.
• Weibo's hot order ends near 300 comments. Newest first goes through all of them.
• About 100 replies per comment at most, as many as Weibo lists.

🧭 Which Chinese social media scraper do you need?

If you want	Use
Weibo posts, comments and the hot search list	This one
Douyin's trending topics	Douyin Hot Search Scraper
Xiaohongshu (RedNote) notes	RedNote Scraper
Xiaohongshu (RedNote) profiles	RedNote Profile Scraper
Bilibili videos, rankings and comments	Bilibili Scraper
Posts on X by keyword	Twitter (X) Search Scraper

❓ Questions people ask

Do I need a Weibo account to scrape Weibo?

No. It reads what Weibo shows anyone who isn't signed in, and needs no login, cookie or API key from you.

Why do I get fewer posts than I see in the Weibo app?

The app is signed in and keeps loading as you scroll. Someone who isn't signed in gets one page per search tab and an account's latest few dozen posts. Open m.weibo.cn in a private browser window: what you see there is what this actor can read. Counts keep moving too, so compare against scrapedAt.

Why are some follower counts round numbers?

Weibo writes big accounts as 44.8万 or 1.6亿 in its lists. authorFollowers turns that into a number with the same rounding, and authorFollowersText keeps Weibo's own text.

Can I call it from code or connect it to an AI assistant?

Yes. The API tab has code for Python, JavaScript and the command line. For Claude, ChatGPT or another MCP client, connect https://mcp.apify.com/?tools=fetch-actor-details,dami_studio/weibo-scraper. Either way it runs on your Apify account at the same price.

Is scraping Weibo legal?

It reads public posts, public comments and the public hot list. Names and comments are still personal data under laws such as GDPR and China's PIPL, and Weibo has its own terms of use, so have a reason and read them. Apify's write-up on scraping and the law is a starting point; we are not lawyers.

🆘 If something breaks

Open the Issues tab on the actor page and send the run ID with the entries you used. The status message and RUN_REPORT say which entry fell short.