Weibo Hot Search Scraper (微博热搜 trending topics API)

Get every Weibo hot-search board (realtime 热搜榜, entertainment 文娱榜, news 要闻榜) as clean rows (rank, topic, heat score, category, label such as new/hot/boom, first-seen time and topic link), plus the top posts of verified accounts for each trending topic. No login, no cookies to paste, $4 per 1,000 results.

{ "mode": "hotSearch", "maxItems": 60 }

One output row (realtime board, rank 1):

{
    "type": "trend",
    "board": "realtime",
    "boardName": "热搜榜",
    "rank": 1,
    "word": "身披旭日旗中国留学生或被开除遣返",
    "heat": 1171091,
    "category": "民生新闻",
    "label": null,
    "isAd": false,
    "isPinned": false,
    "topicUrl": "https://s.weibo.com/weibo?q=%23%E8%BA%AB%E6%8A%AB%E6%97%AD%E6%97%A5%E6%97%97%E4%B8%AD%E5%9B%BD%E7%95%99%E5%AD%A6%E7%94%9F%E6%88%96%E8%A2%AB%E5%BC%80%E9%99%A4%E9%81%A3%E8%BF%94%23",
    "firstSeenAt": "2026-10-19T10:22:22.000Z",
    "minutesOnBoard": 176,
    "snapshotAt": "2026-10-19T13:18:01.174Z",
    "mode": "hotSearch"
}

Compatible with the input of zhorex/weibo-scraper

Switching takes one change: the Actor ID. This Actor accepts the same input fields with the same values: mode, searchQuery, userIds, postIds, maxResults, adFilter (all, organic_only, promo_only), deltaMode and deltaStateKey. It also writes the same output field names next to its own (title, hotValue, labelName, iconDesc, isHot, isNew, isPromoted, url, snapshotAt, firstSeenAt, status, rankDelta, previousRank, hotValueDelta, peakRank, peakHotValue, minutesOnBoard; for posts postId, mid, text, isTruncated, createdAtIso, repostsCount, commentsCount, attitudesCount, authorName, authorVerified, postUrl).

Leader mode	Here
hot_search	Same as hotSearch: every board, not only the realtime one.
hot_search_delta	Same as hotSearch with deltaMode: true (new / rising / falling / steady / dropped).
search	Posts for searchQuery, from verified accounts only, up to maxResults (Weibo shows anonymous visitors about 20 posts per query).
hot_timeline	Posts of Weibo's hot feed, from verified accounts only.
post_comments, post_details, user_posts, user_profile, complaints	Not offered. The run ends at once with a status message explaining why, and charges nothing.

Two differences to know: ads are dropped by default (adFilter: "drop"; pass "all" to keep them flagged), and createdAt is ISO 8601 with Weibo's +08:00 offset (the same value as createdAtIso).

What you get

Board	Weibo name	Rows	Extra fields
realtime	热搜榜	the pinned topic + 50 topics	category, label, Weibo's own first-seen time, paid slot flagged
entertainment	文娱榜	50 topics	category, label
news	要闻榜	50 topics	label

Everything comes from Weibo's public pages and their JSON endpoints, the same ones weibo.com shows to a visitor who is not signed in. A full snapshot of the three boards takes about 10 seconds.

Use cases

• Brands and agencies selling in China: see within the hour when your brand, a competitor or a product category hits the hot search; track how long it stays and how high it climbs.
• Media and newsrooms: monitor what China talks about right now (news, entertainment, society), with the posts of official media accounts on each topic.
• Analysts and researchers: build an hourly time series of Chinese public attention (rank, heat, category) for sentiment, consumer-trend or political-communication studies.
• AI agents: one cheap call returns a structured, deduplicated snapshot of China's trending topics.

Modes

hotSearch (default): trending topics

{ "mode": "hotSearch", "boards": ["realtime", "entertainment", "news"], "maxItems": 60 }

maxItems keeps the top of every board first: 60 rows over 3 boards = the top 20 of each. Leave it empty to get everything (about 150 rows).

topicPosts: what verified accounts post about each trend

{ "mode": "topicPosts", "topTrends": 5, "postsPerTopic": 10 }

Without topics, the Actor takes the current top 5 realtime trends. With topics (e.g. ["#李小冉离婚#", "新能源汽车"]) it searches those. Each post row has the topic, text (full text of long posts), publish time, reposts, comments, likes, author name, verification type (organization = blue V, individual = verified public figure) and link. Set authorFilter: "organizations" to keep media, government and company accounts only.

Only verified accounts are returned, never posts of private individuals. Weibo shows anonymous visitors the first page of results of each search (about 20 posts, hot first); deeper pages need a login, so the Actor stops there.

hot_timeline: Weibo's hot feed

{ "mode": "hot_timeline", "maxItems": 50 }

Build a time series with an hourly schedule

The Actor returns snapshots. To build history, schedule it:

1. Create a task with { "mode": "hotSearch", "deltaMode": true, "deltaStateKey": "hourly" }.
2. In Apify Console go to Schedules → Create new, cron 0 * * * * (every hour), and add the task.
3. Every run stores a full snapshot stamped with snapshotAt. Collect the runs with an integration (Google Sheets, webhook, Make, Zapier) or the API (GET /v2/actor-tasks/{taskId}/runs, then each run's defaultDatasetId items), and group by word and snapshotAt to chart rank and heat over time.
4. With deltaMode, each row also says how it moved since the previous hour: status (new, rising, falling, steady, dropped), previousRank, rankDelta, hotValueDelta, peakRank, peakHotValue. The first run of a delta key is the baseline (every topic new). Use one deltaStateKey per schedule.

Cost of an hourly snapshot of the three boards: about 150 rows × $0.004 = $0.60 per run; the realtime board alone ("boards": ["realtime"]) is about $0.20 per run.

Input

Field	Type	Default	Description
mode	string	hotSearch	hotSearch, topicPosts, hot_timeline, or the leader's hot_search, hot_search_delta, search.
boards	array	all	realtime, entertainment, news.
maxItems	integer	everything	Maximum rows (charged events) in the run. maxResults is an alias.
adFilter	string	drop	drop, keep, only (or organic_only, all, promo_only).
deltaMode	boolean	false	Compare with the previous run (boards) or skip already delivered posts.
deltaStateKey	string	default	Name of the delta stream, stored in your key-value store weibo-trends-state.
topics	array	top trends	topicPosts: topics or keywords to search.
topTrends	integer	5	topicPosts without topics: how many current top trends.
postsPerTopic	integer	10	Maximum posts per topic (in search mode maxResults is used when this is empty).
authorFilter	string	verified	verified or organizations.
fetchFullText	boolean	true	Fetch the full text of long posts (no extra charge).
searchQuery	string		Leader field: query for search mode.
userIds, postIds	array		Leader fields, accepted; their modes are not offered.
proxyConfiguration	object	none	Optional Apify Proxy. Not needed today.

Output fields

Field	Rows	Description
type	all	trend or post (also the charged event).
mode	all	Mode of the run as given in the input.
board, boardName	trend	realtime 热搜榜, entertainment 文娱榜, news 要闻榜.
rank	trend	Position on the board, 1 = top; null for the pinned topic, the paid slot and dropped topics.
word, title	trend	Topic text without # marks.
note	trend	Topic label as Weibo shows it.
isHashtag	trend	true for a #hashtag# topic page.
heat, hotValue	trend	Weibo heat score (热度).
category, categories	trend	Weibo category (民生新闻, 艺人, 剧集, 体育…), as text and as a list.
label	trend	new, hot, boiling, boom, warm, recommended, sponsored, first, exclusive, other or null.
labelName, iconDesc	trend	Badge as shown (新, 热, 沸, 爆…).
isHot, isNew	trend	Badge is 热/沸/爆, badge is 新.
isAd, isPromoted, adType	trend	Paid slot (with adFilter keep/only) and Weibo's ad type.
isPinned	trend	Topic pinned above the realtime board.
topicUrl, url	trend	Weibo search page of the topic.
firstSeenAt, minutesOnBoard	trend	When the topic entered the board (Weibo's own time on the realtime board) and minutes since.
snapshotAt	trend	Time of the snapshot (UTC).
status, previousRank, rankDelta, hotValueDelta, peakRank, peakHotValue	trend	deltaMode only: movement since the previous run.
topic, topicBoard, topicRank	post	Topic or query the post was found for, and where that topic ranked.
postId, mid, bid	post	Weibo post ID and short code.
text, isTruncated	post	Plain text; isTruncated when only the preview was available.
createdAt, createdAtIso	post	Publish time, ISO 8601 with +08:00.
reposts, comments, likes	post	Counts (also as repostsCount, commentsCount, attitudesCount).
authorName, authorId, authorVerified, authorVerifiedType, authorVerifiedReason, authorFollowers	post	Verified author: name, ID, organization or individual, verification text, followers.
images	post	Image URLs.
postUrl	post	Link to the post on weibo.com.
scrapedAt	all	Time of the run (UTC).

Pricing

Pay per result, nothing else: $0.004 per trend row and $0.004 per post row ($4 per 1,000). No platform usage on top, no monthly fee. Rows removed by adFilter or deltaMode are not charged, and unsupported modes charge nothing. Set a maximum cost per run in Apify Console and the Actor stops cleanly when it is reached.

Notes

• No login and no cookie from you: the Actor uses the anonymous visitor session Weibo gives to every browser.
• Weibo's own "topic board" (话题榜) API has not been updated since 2024, so it is left out.
• Times: snapshotAt, firstSeenAt and scrapedAt are UTC; post createdAt keeps Beijing time (+08:00).
• Questions or a missing field? Open an issue on the Actor page; we answer within 24 hours.