Bilibili Search Scraper for Videos and Creators

Search Bilibili (哔哩哔哩 / B站) videos and creator or UP主 accounts by keyword, then export clean, ranked records for content research, creator discovery, trend monitoring, and data analysis. The Actor preserves Bilibili's search ordering, verifies selected video filters, and does not require a Bilibili account or API key.

Choose one result type per run:

• Videos: search by keyword, filter by publication date and duration, sort by relevance, views, publication time, danmaku, or favorites, then enrich every result with public video details and engagement metrics.
• Creators: find Bilibili accounts by keyword and filter or rank them by account type, verification, followers, or user level.

Why use this Actor?

• Two focused search modes: output contains only video records or only creator records in a run.
• Bilibili ordering with reliable filters: sorting stays on Bilibili's search endpoint. Selected publication-date and duration filters are also verified while paging, because Bilibili can ignore those two native parameters when they are combined.
• Detailed video metrics: receive views, likes, coins, favorites, shares, danmaku, comments, tags, category, duration, and author identity when public.
• Useful creator discovery fields: receive followers, video count, level, verification, profile description, creator status, and live-room information when available.
• Clean ranked output: every successful record includes its source keyword and rank.
• Visible empty and failed outcomes: a keyword that returns nothing or cannot be processed creates a structured Dataset status row instead of leaving the Dataset silently empty.
• No account required: the Actor extracts metadata that Bilibili exposes publicly.
• Responsible request pacing: requests are serialized, lightly spaced, and retried only once for temporary network, server, or rate-limit failures.

Input

Common settings

Field	Type	Default	Description
searchQueries	string array, 1–10 items	["人工智能"]	One or more keyword phrases. The limit applies separately to every keyword.
searchType	videos or users	videos	Select one consistent result type for the run. The UI labels users as Creators.
maxResultsPerQuery	integer, 1–50	5	Maximum successful records returned for each keyword. The maximum run size is 500 successful records.

Video-only settings

These values are ignored when searchType is users.

Field	Options	Default	What it does
videoSortBy	relevance, mostViewed, newest, mostDanmaku, mostFavorited	relevance	Uses Bilibili's server-side video result order.
videoUploadDate	any, lastDay, lastWeek, lastSixMonths	any	Limits results by Bilibili publication time and verifies the returned timestamps.
videoDuration	any, under10Minutes, 10To30Minutes, 30To60Minutes, over60Minutes	any	Uses Bilibili's duration buckets and verifies the returned duration.

Creator-only settings

These values are ignored when searchType is videos.

Field	Options	Default	What it does
userType	all, creators, regular, verified	all	Limits results to all users, UP creators, regular users, or verified users.
userSortBy	relevance, mostFollowers, fewestFollowers, highestLevel, lowestLevel	relevance	Uses Bilibili's creator-search ordering.

Quick-start examples

Search videos

{
  "searchQueries": ["人工智能"],
  "searchType": "videos",
  "maxResultsPerQuery": 5,
  "videoSortBy": "mostViewed",
  "videoUploadDate": "lastSixMonths",
  "videoDuration": "over60Minutes"
}

Find creators

{
  "searchQueries": ["人工智能"],
  "searchType": "users",
  "maxResultsPerQuery": 5,
  "userType": "verified",
  "userSortBy": "mostFollowers"
}

How to run it

1. Add one or more values under Search keywords.
2. Select Videos or Creators under Result type.
3. Adjust only the filter section for the selected result type.
4. Click Start.
5. Open the Output tab to review the ranked table or export the Dataset as JSON, CSV, Excel, XML, or JSONL.

Real output samples

The following selected values came from real public Bilibili lookups on September 3, 2026. Live metrics can change after the sample is recorded.

Video record

{
  "recordType": "video",
  "bvid": "BV1qZSLBYEpa",
  "aid": 115626794420560,
  "cid": 34332936080,
  "title": "【整整600集】清华大学196小时讲完的AI人工智能从入门到精通全套教程，全程干货无废话！学完变大佬！这还学不会，我退出IT圈！机器学习-深度学习-opencv",
  "description": "课程涵盖基础概念、算法原理、实践应用，从零开始，深入浅出。通过实例演示，掌握神经网络、决策树、支持向量机等关键技术。适合初学者和进阶者，助您快速提升技能，开启智能时代新篇章。立即观看，开启您的AI学习之旅！",
  "url": "https://www.bilibili.com/video/BV1qZSLBYEpa",
  "thumbnailUrl": "https://i0.hdslb.com/bfs/archive/886e0ea5bd30fdd9ddb66eb5ce108b1a3193d450.jpg",
  "durationSeconds": 195568,
  "durationText": "54:19:28",
  "publishedAt": "2025-11-28T10:20:16Z",
  "categoryId": 231,
  "categoryName": "计算机技术",
  "viewCount": 1023601,
  "likeCount": 24025,
  "coinCount": 19099,
  "favoriteCount": 60224,
  "shareCount": 11470,
  "danmakuCount": 8113,
  "commentCount": 14142,
  "authorMid": 1741962246,
  "authorName": "IT界扛霸子",
  "authorUrl": "https://space.bilibili.com/1741962246",
  "authorAvatarUrl": "https://i0.hdslb.com/bfs/face/330a8ac40ac349d550961219984ba2fc269661ab.jpg",
  "tags": ["IT", "计算机", "AI", "人工智能", "编程", "机器学习"],
  "searchQuery": "人工智能",
  "searchRank": 1,
  "scrapedAt": "2026-09-03T07:18:53.345216Z"
}

Creator record

{
  "recordType": "user",
  "mid": 11515399,
  "name": "极逸人工智能",
  "profileUrl": "https://space.bilibili.com/11515399",
  "avatarUrl": "https://i0.hdslb.com/bfs/face/4a51c264c47b3c759f3efabf51532900e6ac503d.jpg",
  "description": "SOON二测现已圆满结束，感谢全体创作者的鼎力支持与宝贵反馈，我们会继续打磨产品，敬请期待后续版本。",
  "followerCount": 67441,
  "videoCount": 178,
  "level": 6,
  "gender": "male",
  "isCreator": true,
  "isVerified": true,
  "verificationType": "organization",
  "verificationDescription": "极逸人工智能科技有限公司",
  "isLive": true,
  "liveRoomId": 7397397,
  "liveRoomUrl": "https://live.bilibili.com/7397397",
  "searchQuery": "人工智能",
  "searchRank": 2,
  "scrapedAt": "2026-09-03T07:17:02.787817Z"
}

Output fields

recordType: "video"

Group	Fields
IDs and links	bvid, aid, cid, url
Content	title, description, thumbnailUrl, tags
Time	durationSeconds, durationText, publishedAt, publishTimestamp
Category	categoryId, categoryName, categoryV2Id, categoryV2Name
Engagement	viewCount, likeCount, coinCount, favoriteCount, shareCount, danmakuCount, commentCount
Author	authorMid, authorName, authorUrl, authorAvatarUrl
Search context	searchQuery, searchRank, scrapedAt

recordType: "user"

Group	Fields
Identity	mid, name, profileUrl, avatarUrl, description
Account	followerCount, videoCount, level, gender, isCreator
Verification	isVerified, verificationType, verificationDescription
Live room	isLive, liveRoomId, liveRoomUrl
Search context	searchQuery, searchRank, scrapedAt

Fields remain null when Bilibili does not expose a value. For example, a newer category ID can be present while its text label is absent. The Actor does not invent values or convert missing metrics to zero.

Empty-result and error rows

The Dataset receives a searchStatus record when a keyword has no matching public results or a search request fails:

{
  "recordType": "searchStatus",
  "status": "noResults",
  "searchType": "users",
  "searchQuery": "zzzz_no_such_bilibili_user_9f4c2a8d6e1b",
  "searchRank": null,
  "errorCode": "NO_SEARCH_RESULTS",
  "message": "No matching public Bilibili users were found.",
  "scrapedAt": "2026-09-03T07:18:00.635519Z"
}

Video search can also return videoLookupError when a ranked search hit is found but its detail request fails. These status and error rows make failures observable and do not trigger a successful-result billing event.

The default Key-Value Store record named OUTPUT contains totals and per-keyword statuses for the complete run. It also reports requestCount, retryCount, blockedResponseCount, proxyFallbackCount, and videoDetailCacheHits. These operational totals are not added to normal Dataset records.

Run through the Apify API

curl -X POST \
  "https://api.apify.com/v2/acts/arjun_code~bilibili-search-scraper/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "searchQueries": ["人工智能"],
    "searchType": "videos",
    "maxResultsPerQuery": 5,
    "videoSortBy": "relevance",
    "videoUploadDate": "any",
    "videoDuration": "any"
  }'

Pricing

This Actor uses pay-per-event pricing. You pay for successful results, while visible status and lookup-error records do not trigger a result charge.

Event	Free	Bronze	Silver	Gold, Platinum, Diamond
Detailed video result	$8.00 / 1,000	$7.50 / 1,000	$7.00 / 1,000	$6.50 / 1,000
Creator result	$4.00 / 1,000	$3.75 / 1,000	$3.50 / 1,000	$3.25 / 1,000
Actor start	$0.00005 / run	$0.00005 / run	$0.00005 / run	$0.00005 / run

At the Free-plan rate, the default five-result run costs about $0.04005 for videos or $0.02005 for creators. Fifty results cost about $0.40005 or $0.20005 respectively. Platform usage is included in the event price under the current Store configuration.

Billing event names:

• bilibili-video-result — one successfully saved and enriched video record.
• bilibili-creator-result — one successfully saved creator record.
• apify-actor-start — the small run-start charge.
• searchStatus and videoLookupError rows do not trigger a successful-result event; only the run-start charge can remain when no successful result is returned.

The Actor honors the user's maximum event-charge limit and stops adding paid results when that limit is reached. If the same video appears under multiple input keywords, each ranked Dataset result is a separate paid result. Store pricing can change, so the Actor's Pricing tab remains the source of truth.

FAQ

Is a Bilibili account or API key required?

No. This Actor searches metadata that Bilibili exposes publicly. It does not access private, member-only, or account-specific data.

Are video filters applied to creator searches?

No. videoSortBy, videoUploadDate, and videoDuration are ignored for creator searches. Creator filters are likewise ignored for video searches.

Does sorting happen after scraping?

No. The Actor sends the selected sorting to Bilibili and preserves that order in searchRank; it does not locally re-rank results. Date and duration are validated during discovery. When both filters are selected, the Actor keeps the date filter and sorting on Bilibili, then checks duration while paging because Bilibili can otherwise ignore the combined filter parameters.

Why does video search take longer than creator search?

Creator records come directly from search results. Every video result receives one additional detail lookup so the Dataset can include full engagement counts, tags, description, category, and author metadata.

How does the Actor limit request frequency?

Search pages and video-detail calls run sequentially with a small randomized interval. Temporary network and server errors receive at most one delayed retry. HTTP 429 respects Retry-After when available, while explicit risk-control responses stop immediately instead of creating a rapid retry loop. Repeated BV IDs found under different keywords reuse details already fetched during the same run.

The maximum 500-result capacity is intended for controlled batch work, not the fastest interactive response. For faster runs, start with 5–20 results per keyword.

Can one run return videos and creators together?

No. Select one type per run to keep the Dataset schema predictable. Run the Actor twice when you need both datasets.

How many results can one run return?

You can submit up to 10 unique keywords and request up to 50 results per keyword, for a maximum of 500 successful records in one run. The default is 5 results per keyword. Split larger collection jobs into multiple runs to keep runtime and cost predictable.

Does this Actor download video or audio files?

No. It extracts public metadata only.

What is danmaku?

danmakuCount is the public count of time-synchronized on-screen comments associated with a video. commentCount is the count for the conventional comment section.

Support

Use the Actor's Issues tab and include the run ID, a public example keyword, the selected result type and filters, and the relevant status or log message. Never post an Apify token or account credentials.

Legal notice

This Actor is an independent data extraction tool and is not affiliated with, endorsed by, or sponsored by Bilibili. Use it only for lawful purposes and in accordance with applicable terms, privacy rules, and regulations. You are responsible for how you use and store the collected data.