Weibo Post Detail Scraper

✨ What you get

Pull a single Weibo post + its first ~150 comments by postId. Built for marketing monitoring, brand mention triage, and dataset enrichment.

One run returns the detail plus as many comment pages as you request. Pass a postId (and optionally mid, maxPages). Pagination follows the data provider's max_id + max_id_type dual-field protocol (verified 2026-08-20) — no cookies, no auth tokens.

• Actor ID: vgIzEKOSpFKlEDVo5
• Build: 0.1.3
• Source: A hosted Chinese social data API provider (Weibo endpoints)

💼 Use cases

• Reading one post and the conversation under it — postId returns one kind: 'post_detail' record with the full text and counters, plus kind: 'comment' rows carrying floor_number, text and a flat user.
• Walking a deeper thread — maxPages accepts 1 to 10, each page adding about 15 comments, so a full run of 10 pages reaches roughly 150 comments and stops early when the source has no more pages.
• Getting the full post behind a row you already found — this is step ③ of the Weibo set: take a post_id from the hot search or keyword search Actor and turn it into readable text plus replies.
• Keeping the untouched record for audit or enrichment — every row, post and comment alike, preserves the original source object in rawData alongside the normalized fields.

✅ Who this is for

• Brand mention and reputation teams — reading the full text and the replies under a specific post.
• Community managers — working through a thread to see who said what, in order.
• Researchers and analysts — enriching a dataset of post ids with complete post and comment rows.
• Cross-border sellers and agencies — needing the Chinese-language conversation around a post or a product mention.

❌ Who this is NOT for

• You need the whole comment thread — maxPages caps at 10 and each page is about 15 comments, so roughly 150 is the ceiling and pagination stops early when the source runs out. The post's own comments_count can be far higher than what you fetched; treat the comments you received as the top of the thread, not the full discussion.
• You need a nested conversation tree — comments come back as a flat list of kind: 'comment' rows with a floor_number, so parent and child replies are not handed to you as a structure you can walk.
• You need numeric audience sizes for the authors — followers_count is a formatted display string such as "2.3万" on the user profile, not a number. Parse it before you rank authors or filter by audience.
• You are starting from a keyword or a trend rather than a post id — postId is required and this Actor does not search. Finding posts is step ① or ② of the Weibo set; this is the leaf that opens one known post.
• You need to pull many posts at once, or you are on a free account — one run is one post, and free users are limited to 5 runs per day. A backfill over hundreds of post_id values needs a paid tier and many scheduled runs.

🚀 Quick start

{
  "postId": "5328347665010605",
  "mid": "5328347665010605",
  "maxPages": 1
}

Output dataset mixes post detail and comments, each tagged with kind:

[
  {
    "kind": "post_detail",
    "post_id": "5328347665010605",
    "mid": "5328347665010605",
    "created_at": "Wed Aug 05 00:22:22 +0800 2026",
    "text": "现在拍戏都这么拼吗 [允悲] 女演员工伤 偷胯婆的微博视频",
    "text_length": 57,
    "source": "iPhone 17 Pro Max",
    "reposts_count": 142,
    "comments_count": 523,
    "attitudes_count": 1820,
    "user": { "user_id": 7491687377, "screen_name": "偷胯婆", "verified": true, "followers_count": "2.3万" },
    "rawData": { /* full Data provider post object */ }
  },
  {
    "kind": "comment",
    "post_id": "5328347665010605",
    "comment_id": "5010203873098001",
    "floor_number": 1,
    "created_at": "Wed Aug 05 00:30:11 +0800 2026",
    "text": "心疼姐姐 [泪]",
    "user": { "user_id": 1234567890, "screen_name": "路人甲" },
    "rawData": { /* full Data provider comment object */ }
  }
]

📥 Input parameters

Field	Required	Default	Notes
postId	yes	—	Numeric Weibo post id (string of digits, ≥6 chars).
mid	no	falls back to postId	Numeric Weibo mid; required by Data provider for fetch_post_comments.
maxPages	no	1	1..10. Each page returns ~15 comments; maxPages=10 ≈ 150 comments total. Pagination stops early when Data provider has no more data.

📤 Output

The Actor pushes one kind: 'post_detail' item + N kind: 'comment' items to the default dataset and writes a SUMMARY record to the default key-value store with run details (postId, mid, pagesFetched, requests, uniqueComments, totalCommentsFromData provider, nextMaxId, nextMaxIdType, free-run counter, etc.).

💰 Billing

Free users are limited to 5 runs per day. Paid users (APIFY_USER_IS_PAYING=1) have unlimited runs. The Actor itself adds no markup — you pay the data provider provider per request plus the standard Apify compute unit cost.

📚 Resources

Data source: a hosted Chinese social data API provider (Weibo endpoints). The provider charges per Data provider request; this Actor adds no markup.

MIT

📜 Notes & caveats

Use it together with the other Weibo Actors

This Actor is the step ③ leaf node of the 3-Actor Weibo MVP-2 set. Use it together with:

Where in the pipeline	Actor	Console	Use it to
before / step ①	Weibo Hot Search Scraper	Console	step ① — fetch the realtime hot list, get post_id of each trend
before / step ②	Weibo Keyword Search Scraper	Console	step ② — search posts by keyword, get post_id of each result
after / step ③	Weibo Post Detail Scraper (you are here)	(step ③ — leaf node)	step ③ — pass in one post_id from above, pull full post text + up to ~150 comments

You are here: Weibo Post Detail Scraper — run this on its own for the data provider use case described above.

Why this Actor exists

Apify has ~50,000 Actors and almost zero Weibo scrapers. This Actor is part of a planned Weibo MVP-2 set (5 Actors covering hot search, keyword search, user profile, post detail + comments, and channel feed) that targets the gap.