Weibo (微博) Scraper

WEIBO (微博) SCRAPER  •  WEIBO API See who amplified it, not just how many. Ten modes on one clean shape: keyword search, user timelines, post details, comments, profiles and the hot-search board — plus the lists behind the numbers that count-only tools never open: who reposted, who liked, a user's fans and who they follow, each returned as a full profile. Keyword-search filters are free. No Weibo account, no login, no cookies, on any mode. PAY PER RESULT, FAILED LOOKUPS ARE FREE  10 MODES, UP TO 5,000 RESULTS PER TARGET

Copy to your AI assistant

Paste this into ChatGPT, Claude, Cursor, or any LLM to start using this Actor right away.

atomus/weibo-scraper is an Apify Actor that returns public Weibo (微博) data as structured JSON in ten modes chosen by the required searchType input: "search" (keyword search, from keywords), "timeline" (a user's posts, from userUrls), "post-detail" (one post by URL or id, from postUrls), "comments", "reposts" (the accounts that reposted a post), "likes" (the accounts that liked it), "profile" (a full user profile), "fans" (a user's followers), "following" (the accounts a user follows) and "trending" (the hot-search board). Use it whenever someone needs Weibo posts, comments, profiles, trends, or the virality and audience graph behind an engagement number, without cookies, a login, or a browser - China market research, KOL vetting, brand monitoring, or trend detection. Run it with curl: curl -X POST "https://api.apify.com/v2/acts/atomus~weibo-scraper/run-sync-get-dataset-items?token=APIFY_TOKEN" -H "Content-Type: application/json" -d '{"searchType":"search","keywords":["新能源汽车"],"timeFilter":"7d","searchScope":"hot","maxItems":50}'. Or in Python: ApifyClient("APIFY_TOKEN").actor("atomus/weibo-scraper").call(run_input={"searchType":"reposts","postUrls":["https://weibo.com/1896820725/RcBE4tNYz"],"maxItems":200}) then client.dataset(run["defaultDatasetId"]).list_items().items. Inputs: searchType (required, default search), keywords (string[], search mode), postUrls (string[], accepts a weibo.com/uid/mblogid URL, an m.weibo.cn/detail/id link, a bare base62 mblogid or a numeric mid), userUrls (string[], a weibo.com/u/uid URL or a bare numeric user id), maxItems (int, default 20, max 5000 per keyword or target), includeRankings (bool, trending mode only), and the FREE keyword-search filters timeFilter (all, 1d, 7d, 30d, 180d, 365d), searchScope (all, hot, original, verified, media, viewpoint) and mediaType (all, pic, video, music, link). Every row carries searchType so mixed datasets stay sortable; a post row has mid, mblogid, text, created_at, source, region, reposts_count, comments_count, attitudes_count, topics, images, videos, url and a nested user object; a comment row adds like_counts, floor_number, liked_by_author and a replies preview; a profile row has followers_count, friends_count, statuses_count, verified_reason, location and description; a trend row has rank, keyword, heat and tag. Billing is per result, per mode, with a per-target minimum of 2 results for comments, reposts, likes, fans, following and trending and 1 for search, timeline, post detail and profile; a request that ERRORS is free, a request that succeeds and finds nothing still bills that minimum, and Gold and higher Apify plans pay 20 percent less. Full input schema, every enum and default, and the complete output field list: GET https://api.apify.com/v2/acts/atomus~weibo-scraper/build/default

---

Weibo MCP Server: use these Actors from ChatGPT, Claude or Cursor

Point your AI assistant at Atomus and it can read the social platforms on its own: Weibo, RedNote, Douyin, Bilibili, TikTok and X. No glue code, no scraping logic in your prompts.

{
  "mcpServers": {
    "atomus": {
      "url": "https://mcp.apify.com?tools=atomus/weibo-scraper,atomus/xiaohongshu-scraper,atomus/douyin-scraper,atomus/bilibili-scraper,atomus/tiktok-scraper,atomus/tiktok-comments-scraper,atomus/twitter-scraper",
      "headers": { "Authorization": "Bearer YOUR_APIFY_TOKEN" }
    }
  }
}

Then ask, in plain language:

"Find the most-reposted Weibo posts about 新能源汽车 from the last week, then for the top one list every verified account that reposted it."

That one sentence uses this Actor in two modes. Pinning the tools= list is what keeps your assistant on these Actors instead of reaching for whatever scraper it finds first.

---

How to scrape Weibo

searchType picks the mode, and each mode reads one input field: keywords, postUrls or userUrls. Fill only that one.

Keyword search, narrowed for free

{
  "searchType": "search",
  "keywords": ["新能源汽车", "科技"],
  "timeFilter": "7d",
  "searchScope": "hot",
  "mediaType": "video",
  "maxItems": 100
}

timeFilter, searchScope and mediaType cost nothing and are applied by Weibo before results come back, so a narrower search is also a cheaper one.

A user's timeline

{ "searchType": "timeline", "userUrls": ["https://weibo.com/u/1896820725", "2803301701"], "maxItems": 50 }

Returns that account's own posts. A full profile URL and a bare numeric user id both work.

Who reposted a post, the virality graph

{ "searchType": "reposts", "postUrls": ["https://weibo.com/1896820725/RcBE4tNYz"], "maxItems": 200 }

Not the repost count: each reposting post, with the account behind it. This is what tells you which accounts actually carried a story.

Who liked a post, the audience graph

{ "searchType": "likes", "postUrls": ["https://m.weibo.cn/detail/5328843155702975"], "maxItems": 500 }

Every account that liked the post, as a profile with followers, verification, bio and location.

A KOL's fans, or who they follow

{ "searchType": "fans", "userUrls": ["2803301701"], "maxItems": 1000 }

Swap fans for following to walk the other direction. Some accounts keep these lists private, and then the run returns nothing for that target.

Every comment on a post

{ "searchType": "comments", "postUrls": ["RcBE4tNYz"], "maxItems": 300 }

Comments carry the like count, the floor number, the commenter's region, whether the post author liked the comment, and a preview of the replies underneath.

One post, or one profile, in full

{ "searchType": "post-detail", "postUrls": ["5328843155702975"] }

post-detail and profile are one billed record per target, whatever maxItems says.

Today's hot-search board

{ "searchType": "trending", "includeRankings": true, "maxItems": 50 }

The Weibo hot-search board with rank, heat and tag. includeRankings also returns the social ranking board, billed as extra trend rows.

---

Input

Parameter	Type	Required	Default	Used by	Description
searchType	string	✅ Yes	search	—	search, timeline, post-detail, comments, reposts, likes, profile, fans, following, trending.
keywords	string[]	per mode	(none)	search	One or more search terms, e.g. 科技.
postUrls	string[]	per mode	(none)	post-detail, comments, reposts, likes	Post URLs or IDs.
userUrls	string[]	per mode	(none)	timeline, profile, fans, following	Profile URLs or numeric user IDs.
maxItems	integer	No	20	every list mode	1 to 5,000 per keyword or target. Your main cost lever.
includeRankings	boolean	No	false	trending	Also return the social ranking board (billed as extra trend rows).

Free keyword-search filters, no extra charge: timeFilter (all, 1d, 7d, 30d, 180d, 365d), searchScope (all, hot, original, verified, media, viewpoint) and mediaType (all, pic, video, music, link). Weibo applies them server-side, so you are billed only for what comes back.

Which URLs and IDs work

Posts:
https://weibo.com/1896820725/RcBE4tNYz    uid / mblogid
https://m.weibo.cn/detail/5328843155702975
RcBE4tNYz                                 bare base62 mblogid
5328843155702975                          bare numeric mid

Users:
https://weibo.com/u/2803301701
https://weibo.com/2803301701
2803301701                                bare user id

If the field a mode reads is empty, the run stops immediately with an explanation in the dataset and charges nothing.

---

What data does the Weibo Scraper return?

One flat row per result. Every row carries searchType, so a dataset built from several runs stays sortable by mode.

Group	Fields
Post (search, timeline, post-detail, reposts)	mid · mblogid · text · created_at · source · region · topics[] · is_retweet · is_long_text · url
Post engagement	reposts_count · comments_count · attitudes_count (likes)
Post media	pic_num · pic_ids[] · images[] (full HTTPS URLs) · videos[] · has_image · has_video · video_page_url
Comment (comments)	id · text · like_counts · floor_number · created_at · source · region · liked_by_author · reply_count · replies[] (inline preview) · post_id
User, on every row	user.uid · screen_name · avatar · verified · verified_reason · verified_type · member_rank · svip · followers_count · location · description · profile_url
Profile rows (profile)	the user fields above, plus friends_count (following) · statuses_count · gender · external_url · cover_image · top_user · is_star
Fans / following / likes rows	the user fields above, as a top-level record, plus of_user (fans, following) or post_id (likes)
Trend (trending)	rank · keyword · heat · tag · is_top · url · board (hot_search or social_ranking)

Example row

A keyword-search result, trimmed to show the shape and the nesting.

{
  "searchType": "search",
  "keyword": "科技",
  "mid": "5328843155702975",
  "mblogid": "RcBE4tNYz",
  "text": "科技板块目前看反弹也不会一帆风顺 …",
  "created_at": "Wed Aug 06 09:11:00 +0800 2026",
  "source": "微博网页版",
  "region": "发布于 天津",
  "reposts_count": 12,
  "comments_count": 34,
  "attitudes_count": 210,
  "topics": ["科技股"],
  "is_retweet": false,
  "images": ["https://wx1.sinaimg.cn/large/…jpg"],
  "videos": [],
  "url": "https://weibo.com/1896820725/RcBE4tNYz",
  "user": {
    "uid": "1896820725",
    "screen_name": "天津股侠",
    "verified": true,
    "verified_reason": "财经博主",
    "followers_count": 812340,
    "location": "天津",
    "profile_url": "https://weibo.com/u/1896820725"
  }
}

Two platform limits, stated plainly

Some accounts keep their fans and following lists private. When a list is not public, that target comes back empty — and because the lookup still happened, it bills the 2-result minimum for that mode.

Weibo paginates each mode at its own page size, and the Actor follows pagination as far as Weibo exposes it, up to your maxItems. A very deep target can therefore stop short of the number you asked for.

---

How much does it cost to scrape Weibo?

Pay per result, as a per-event charge. Each mode is its own event.

Mode → event	Free, Bronze, Silver	Gold+
Keyword search, user timeline → post-scraped	$20.00 / 1,000	$16.00 / 1,000
Post detail → post-detail-scraped	$40.00 / 1,000	$32.00 / 1,000
Profile → profile-scraped	$40.00 / 1,000	$32.00 / 1,000
Reposts → repost-scraped	$12.00 / 1,000	$9.60 / 1,000
Fans → fan-scraped	$12.00 / 1,000	$9.60 / 1,000
Comments → comment-scraped	$10.00 / 1,000	$8.00 / 1,000
Likes → like-scraped	$10.00 / 1,000	$8.00 / 1,000
Following → following-scraped	$10.00 / 1,000	$8.00 / 1,000
Trending → trend-scraped	$10.00 / 1,000	$8.00 / 1,000

Gold and higher Apify plans pay 20% less on every event.

A request that fails is free. If the lookup errors, nothing is charged for that target.

A request that succeeds but finds nothing still bills that mode's minimum: 2 results for comments, reposts, likes, fans, following and trending, and 1 for keyword search, user timeline, post detail and profile. One request is one lookup, and Weibo charges us for it whether it comes back with 50 rows, 1, or none — the minimum covers the request itself, which is what keeps thin targets worth running instead of refused outright.

Search filters are free, and they run before results are counted, so a tighter query is a cheaper one.

Free plan: 10 posts per calendar month, plus 1 result of every other type, sampled from a single target per run, so you can see every record shape before paying. The counters reset on the 1st.

---

What do people use the Weibo Scraper for?

• Virality analysis: pull the full repost list of a post to map how it spread and which verified accounts drove it, instead of guessing from a number.
• Audience research: export a KOL's fans or following list and profile who actually follows them.
• Influencer vetting: check a profile's real follower count, verification reason and recent posts before signing a partnership.
• Brand and campaign monitoring: search a brand or hashtag on a schedule and track posts, comments and sentiment over time.
• China market research: watch a category with the free time and media filters, at a cost you can bound with maxItems.
• Trend detection: read the hot-search board for topics breaking in the China market before they reach Western coverage.
• Crisis and PR monitoring: collect the public statements and the reactions around an event or a public figure.
• AI agent context: give an agent live Chinese-language public discussion instead of its training cutoff.
• AI and research datasets: one consistent post, comment and profile shape across all ten modes.

---

Weibo Scraper vs cookie-based tools vs the official Weibo API

	This Weibo Scraper	Cookie-based scrapers	Official Weibo API
Weibo account / cookies	Not needed, on any mode	Session cookie required, especially for user data	Registered developer app
Account / ban risk	None (no account used)	High (your account can be flagged)	None
Setup	Paste a keyword, post URL or profile URL	Extract and paste your cookie	App registration + review
Reposts, likes, fans, following as full lists	Yes, each as a full profile	Counts only, or partial	Gated
Timeline, fans, following without a login	Yes	Usually needs the cookie	Gated
Pricing	Pay per result (from $16.00 / 1,000 posts on Gold)	Subscription + your own account	App approval, heavy rate limits

---

🏆 Top China Scrapers

	RedNote Scraper Notes, user profiles and comments from 小红书
	Douyin Scraper Creator profiles, videos and comments from 抖音
	Weibo Scraper  YOU ARE HERE
	Bilibili Scraper Videos, creators and stats from 哔哩哔哩

---

FAQ

Do I need a Weibo account or cookies?

No. The Actor reads public Weibo data and handles the request signing itself, on every mode — including user timelines, fans and following, which many Weibo scrapers can only return when you supply a logged-in cookie. There is nothing to log into and no account to put at risk.

What can I get here that count-only Weibo scrapers can't?

The lists behind the numbers: the actual accounts that reposted or liked a post, and a user's full fans and following lists, each returned as a profile with followers, verification, bio and location.

Can I use this from ChatGPT or Claude?

Yes, two ways. Paste the "Copy to your AI assistant" block above into any LLM and it will write the call for you. Or connect the Weibo MCP server config above, and every Atomus Actor becomes a native tool your assistant can call on its own, including chaining several in one request.

How do I pass a post or a user?

A post accepts a weibo.com/<uid>/<mblogid> URL, an m.weibo.cn/detail/<mid> link, a bare base62 mblogid, or a numeric mid. A user accepts a weibo.com/u/<uid> URL or a bare numeric user id.

Can I filter a keyword search by date or media, and does it cost more?

Yes, and no. timeFilter, searchScope and mediaType are free and are applied by Weibo before the results are counted, so narrowing a search lowers the bill.

Why did a fans or following run return nothing, and was I charged?

Some accounts keep those lists private. The lookup still ran, so it bills that mode's 2-result minimum. Only a request that errors is free.

How much does it cost?

$20 per 1,000 posts from keyword search or a timeline, $40 per 1,000 post details or profiles, $12 per 1,000 reposts or fans, $10 per 1,000 comments, likes, following rows or trends. Gold and higher plans pay 20% less. Free Apify plans include 10 posts and 1 of every other result type per month.

How many results can I get per target?

Up to 5,000 per keyword or target. The Actor follows Weibo's pagination as far as the platform exposes it, so a thin target can return fewer.

Is there an official Weibo API?

Yes, but Weibo's open platform requires a registered developer app, gates most endpoints and is heavily rate-limited. This Actor is the public-data alternative: no app registration, no keys, structured JSON, pay per result.

Is it legal to scrape Weibo?

This Actor collects only publicly available data. You are responsible for using the output in line with applicable laws (GDPR/CCPA), Weibo's terms, and your own compliance requirements. It is an independent tool, not affiliated with Sina Weibo.

Can I run this on a schedule?

Yes. Apify Schedules run the Actor on a cron interval and webhooks push each finished run into your systems — which is how a keyword search becomes a brand monitor.

---

微博采集器中文说明

按关键词或链接抓取公开的微博数据：关键词搜索、用户时间线、微博详情、评论、用户资料、热搜榜，以及大多数工具只给一个数字的传播与受众关系图：谁转发了、谁点赞了、用户的粉丝与关注列表，每一条都是完整的用户资料。所有模式均无需 Cookie、无需登录、无需微博账号。

十种模式，各返回一种记录

模式 (searchType)	返回内容
search 关键词搜索	微博：正文、转发/评论/点赞数、来源、发布地、话题、图片与视频、作者
timeline 用户时间线	该账号自己发布的微博，字段同上
post-detail 微博详情	单条微博的完整字段（每个目标计一条）
comments 评论	评论正文、点赞数、楼层、发布地、是否被博主点赞、回复预览
reposts ⭐ 转发列表	每一条转发微博及其背后的账号（传播图谱）
likes ⭐ 点赞列表	点赞该微博的账号，每个都是完整资料（受众图谱）
profile 用户资料	粉丝数、关注数、微博数、认证与认证理由、简介、地区（每个目标计一条）
fans ⭐ 粉丝列表	该用户的粉丝，每个都是完整资料
following ⭐ 关注列表	该用户关注的账号，每个都是完整资料
trending 热搜榜	热搜条目：排名、关键词、热度、标签（可选社会榜单）

⭐ = 只给数字的微博采集器无法提供的关系图谱数据。

如何采集

searchType 决定模式，每种模式只读取一个输入字段：keywords、postUrls 或 userUrls，只填这一个即可。

关键词搜索（筛选项免费）：

{
  "searchType": "search",
  "keywords": ["新能源汽车", "科技"],
  "timeFilter": "7d",
  "searchScope": "hot",
  "mediaType": "video",
  "maxItems": 100
}

查看谁转发了某条微博（传播图谱）：

{ "searchType": "reposts", "postUrls": ["https://weibo.com/1896820725/RcBE4tNYz"], "maxItems": 200 }

导出某位达人的粉丝列表（受众图谱）：

{ "searchType": "fans", "userUrls": ["2803301701"], "maxItems": 1000 }

输入参数

参数	类型	必填	默认值	适用模式	说明
searchType	string	✅ 是	search	—	上表中的十种模式之一。
keywords	string[]	视模式	无	关键词搜索	一个或多个搜索词，例如 科技。
postUrls	string[]	视模式	无	详情、评论、转发、点赞	微博链接或 ID。
userUrls	string[]	视模式	无	时间线、资料、粉丝、关注	主页链接或数字用户 ID。
maxItems	integer	否	20	所有列表模式	每个关键词或目标 1 至 5,000 条，是控制费用的主要开关。
includeRankings	boolean	否	false	热搜	额外返回社会榜单（按热搜条目计费）。

免费筛选项（仅关键词搜索，不额外计费）： timeFilter（all、1d、7d、30d、180d、365d）、searchScope（all、hot、original、verified、media、viewpoint）、mediaType（all、pic、video、music、link）。由微博服务端过滤，因此只对返回的结果计费。

支持的链接与 ID： 微博链接 https://weibo.com/<uid>/<mblogid>、https://m.weibo.cn/detail/<mid>、纯 base62 mblogid、纯数字 mid；用户主页 https://weibo.com/u/<uid> 或纯数字用户 ID。

输出字段

每条结果为一行扁平 JSON，并且每一行都带有 searchType，多次运行汇总到同一数据集也能按模式区分。

• 微博：mid、mblogid、text、created_at、source、region、topics[]、is_retweet、is_long_text、url
• 互动数：reposts_count、comments_count、attitudes_count（点赞）
• 媒体：pic_num、pic_ids[]、images[]、videos[]、has_image、has_video、video_page_url
• 评论：id、text、like_counts、floor_number、created_at、source、region、liked_by_author、reply_count、replies[]、post_id
• 用户（每行都有）：user.uid、screen_name、avatar、verified、verified_reason、verified_type、member_rank、svip、followers_count、location、description、profile_url
• 用户资料行：在上述字段外，另有 friends_count（关注数）、statuses_count、gender、external_url、cover_image、top_user、is_star
• 粉丝／关注／点赞行：上述用户字段作为顶层记录，另加 of_user（粉丝、关注）或 post_id（点赞）
• 热搜：rank、keyword、heat、tag、is_top、url、board（hot_search 或 social_ranking）

费用

按结果计费，每种模式是独立的付费事件。

模式 → 事件	Free / Bronze / Silver	Gold 及以上
关键词搜索、用户时间线 → post-scraped	$20.00 / 1,000	$16.00 / 1,000
微博详情 → post-detail-scraped	$40.00 / 1,000	$32.00 / 1,000
用户资料 → profile-scraped	$40.00 / 1,000	$32.00 / 1,000
转发列表 → repost-scraped	$12.00 / 1,000	$9.60 / 1,000
粉丝列表 → fan-scraped	$12.00 / 1,000	$9.60 / 1,000
评论 → comment-scraped	$10.00 / 1,000	$8.00 / 1,000
点赞列表 → like-scraped	$10.00 / 1,000	$8.00 / 1,000
关注列表 → following-scraped	$10.00 / 1,000	$8.00 / 1,000
热搜 → trend-scraped	$10.00 / 1,000	$8.00 / 1,000

• Gold 及以上 Apify 套餐全部 8 折。
• 请求出错免费，该目标不计任何费用。
• 请求成功但无结果，仍按该模式的最低条数计费：评论、转发、点赞、粉丝、关注、热搜为 2 条，关键词搜索、用户时间线、微博详情、用户资料为 1 条。一次请求就是一次查询，无论返回 50 条、1 条还是 0 条，微博对我们的收费都一样；最低计费覆盖的是这次请求本身，这样结果很少的目标依然值得抓取，而不是被直接拒绝。
• 筛选项免费，且在计费之前生效，所以查询越精准费用越低。
• 免费套餐：每个自然月可采集 10 条微博，外加其他每种结果各 1 条，且每次运行只取一个目标，足以在付费前查看所有字段结构。计数在每月 1 日重置。

使用场景

• 传播分析：拉取整份转发列表，还原一条微博如何扩散、由哪些认证账号推动。
• 受众研究：导出达人的粉丝或关注列表，刻画其真实受众。
• 达人筛选：合作前核实真实粉丝数、认证理由与近期内容。
• 品牌与舆情监测：定时搜索品牌词或话题，跟踪微博、评论与风向变化。
• 趋势发现：读取热搜榜，第一时间捕捉正在发酵的话题。
• AI 智能体上下文：把结构化的微博、评论与受众数据交给大模型做摘要或分类。

常见问题

需要微博账号或 Cookie 吗？ 不需要。所有模式都只读取公开数据，包括用户时间线、粉丝与关注列表——这些内容在很多微博采集器上必须提供登录 Cookie 才能获取。没有账号被限制的风险。

和只给数字的采集器相比，多了什么？ 数字背后的名单：真正转发、点赞该微博的账号，以及用户完整的粉丝与关注列表，每一条都带粉丝数、认证、简介和地区。

粉丝／关注列表为什么是空的？会计费吗？ 部分账号不公开这两个列表。此时查询依然发生了，因此按该模式的最低 2 条计费；只有请求出错才完全免费。

每个目标最多能取多少条？ 每个关键词或目标最多 5,000 条。采集器会按微博开放的分页深度尽可能翻页，因此内容较少的目标可能返回不足。

微博有官方 API 吗？ 有，但微博开放平台需要注册开发者应用、审核，多数接口受限且有严格频率限制。本采集器是面向公开数据的替代方案：无需注册应用、无需密钥，直接返回结构化 JSON，按结果计费。

可以在 ChatGPT 或 Claude 里使用吗？ 可以。把上面的「Copy to your AI assistant」整段粘贴进任意大模型即可，或直接配置上文的 Weibo MCP server，让助手把每个 Atomus Actor 当作原生工具调用。

---

All Atomus scrapers

2.4M+ RESULTS DELIVERED

LinkedIn	Profile · Posts · Reactions · Comments · Company · Employees
Lead gen	Leads Finder
Google Maps	Places & local businesses
TikTok	Videos · Comments
X (Twitter)	Tweets & profiles
RedNote 小红书	Notes, users & comments
Douyin 抖音	Profiles, videos & comments
Weibo 微博	Posts & profiles
Bilibili 哔哩哔哩	Videos & creators

---

Support

Hey, I'm Chico, founder of Atomus. I built this Actor and I answer the messages about it. Something broke? A field you need isn't there? Not sure it fits what you're doing? Send me a message, most answers come the same day. 💬  DM me on LinkedIn    or hello@dendelabs.com

---

⚠️ Disclaimer

This Actor is an independent tool and is not affiliated with, endorsed by, or sponsored by Sina Corporation or Weibo Corporation. Weibo® and 微博® are trademarks of their respective owners. All trademarks are property of their respective owners.

Use the data extracted by this Actor in compliance with applicable data protection laws (GDPR, CCPA) and Weibo's terms of service. Do not use it for spam, harassment, or unlawful purposes.

本采集器为独立工具，与新浪公司及微博公司无任何隶属、认可或赞助关系。微博® 为其权利人的注册商标。请在符合适用数据保护法律（GDPR、CCPA）及微博服务条款的前提下使用所采集的数据，不得用于垃圾信息、骚扰或任何违法用途。