Bilibili Scraper – Videos, Search, Rankings, Comments & Danmaku

Bilibili scraper for public data, no login: popular (热门) and ranking (排行榜) videos, keyword search, video details with views, likes, coins and favorites, comments, danmaku (弹幕 bullet comments) and uploader (UP主) search. Use it instead of a general web crawler when you need clean Bilibili JSON with every mode in one Actor at $2 per 1,000 videos.

What data do you get?

• 🎬 Videos – BV id, av id, title, description, cover, duration, publish date, category, tags, uploader name and id, and statistics: views, likes, coins, favorites, shares, danmaku count and comment count
• 🔥 Popular list – Bilibili's current 热门 feed (up to ~500 videos) with the recommendation reason (e.g. "百万播放")
• 🏆 Rankings – top-100 排行榜 for all of Bilibili or one zone (gaming, music, tech, knowledge, food, dance…), overall / original / rookie
• 🔎 Search – videos for any Chinese or English keyword, sorted by relevance, views, newest, danmaku or favorites (up to 1,000 per keyword)
• 💬 Comments – pinned and hot comments plus their replies: text, likes, reply count, date, whether the uploader wrote it
• 📜 Danmaku – bullet comments with the exact video timestamp, text, display mode (scroll / top / bottom), color and send date
• 👤 Uploaders – UP主 search: name, UID, followers, number of videos, level, verification and latest videos

All items are flat JSON with consistent field names. Export as JSON, CSV, Excel, XML or HTML, or use the Apify API.

Use cases

• Trend & content research – what is hot in China right now, per category, every day.
• Brand and game monitoring – track videos and audience reactions for your game, product or brand keyword.
• Influencer (KOL) discovery – find UP主 by topic and compare followers and engagement (coins, favorites, likes per view).
• Audience sentiment & NLP – danmaku and comments are a uniquely rich, time-coded reaction dataset for sentiment analysis and LLM training/evaluation.
• Market entry for China – understand which formats, topics and creators win on Bilibili before you invest.

How to use

1. Pick What to scrape: Popular, Ranking, Search, Specific videos or Search uploaders.
2. Fill in keywords (search modes) or video links / BV ids (videos mode).
3. Optionally switch on Include comments and/or Include danmaku.
4. Click Start. The default input ({} = 20 popular videos) finishes in about 15 seconds.

Input example

{
    "mode": "search",
    "searchQueries": ["原神"],
    "searchOrder": "totalrank",
    "maxItems": 5,
    "includeComments": true,
    "maxCommentsPerVideo": 100,
    "includeDanmaku": true,
    "maxDanmakuPerVideo": 0
}

Field	Type	Description
mode	enum	popular (default), ranking, search, videos, searchUploaders
searchQueries	array	Keywords for search / searchUploaders
videoUrls	array	Video links, BV… ids or av… numbers for videos
rankingCategory	enum	Zone id for ranking (0 = all, 4 = gaming, 188 = tech, …)
rankingType	enum	all, origin, rookie
searchOrder	enum	totalrank, click, pubdate, dm, stow
maxItems	integer	Max videos (or uploaders) per run / per keyword. Default 20
includeTags	boolean	Load each video's tag list (free)
includeComments / maxCommentsPerVideo	boolean / integer	Comments per video (default 20)
includeDanmaku / maxDanmakuPerVideo	boolean / integer	Danmaku per video (default 200, 0 = all)
proxyConfiguration	object	Optional proxy (see FAQ)

Output sample

Real items from a run on 2026-10-08 (shortened):

{
    "type": "video",
    "source": "search",
    "rank": 1,
    "searchQuery": "原神",
    "bvid": "BV1F8aJ6TEKz",
    "aid": "117353505429329",
    "cid": "42304735801",
    "url": "https://www.bilibili.com/video/BV1F8aJ6TEKz",
    "title": "《原神》剧情PV-「往日」",
    "coverUrl": "https://i1.hdslb.com/bfs/archive/489850205a82aba0b3c246f69b124fdc6cfe35d3.jpg",
    "durationSec": 1402,
    "publishedAt": "2026-10-07T04:00:00.000Z",
    "category": "手机游戏",
    "tags": ["原神", "miHoYo", "米哈游"],
    "uploaderName": "原神",
    "uploaderId": "401742377",
    "uploaderUrl": "https://space.bilibili.com/401742377",
    "views": 495644,
    "likes": 67441,
    "coins": 11878,
    "favorites": 9550,
    "shares": 2959,
    "danmakuCount": 4861,
    "commentCount": 9237,
    "partsCount": 4
}

{
    "type": "comment",
    "bvid": "BV1F8aJ6TEKz",
    "commentId": "316322760481",
    "rootCommentId": null,
    "isReply": false,
    "isPinned": false,
    "isByUploader": false,
    "text": "海包海包帮我生成停止圣土化的方法",
    "likes": 3861,
    "replyCount": 18,
    "postedAt": "2026-10-07T04:10:23.000Z"
}

{
    "type": "danmaku",
    "bvid": "BV1F8aJ6TEKz",
    "cid": "42304735801",
    "danmakuId": "2216727619263765248",
    "videoTimeSec": 0,
    "text": "往日种种…",
    "mode": "top",
    "fontSize": 25,
    "color": "#ffffff",
    "sentAt": "2026-10-07T04:02:04.000Z"
}

Other item types: uploader (uploaderId, uploaderName, followers, videosCount, level, verified, latestVideos) and error (free – errorCode is INVALID_INPUT, NOT_FOUND, NO_RESULTS, BLOCKED, NO_CID or SCRAPE_FAILED). The dataset has views: Overview, Comments, Danmaku, Uploaders & errors.

Pricing

Pay per event – you pay only for results. Error rows are free.

Event	Price	Per 1,000
video – one video row	$0.002	$2.00
comment – one comment	$0.001	$1.00
danmaku – one bullet comment	$0.0005	$0.50
uploader – one uploader (UP主) row	$0.002	$2.00

Examples: 100 ranking videos = $0.20. 5 search videos + 194 comments + 2,030 danmaku (real run above) = $0.01 + $0.194 + $1.015 = $1.22. Set Max cost per run in the run options – the Actor stops cleanly at your budget and never pushes unpaid items.

Use with AI agents (MCP)

The Actor works as a tool in Apify's MCP server, so Claude, Cursor and other MCP clients can call it. Add the Actor in your MCP client by its Store name and send input like:

{
    "mode": "ranking",
    "rankingCategory": "188",
    "maxItems": 10,
    "includeTags": true
}

The agent gets back flat rows with title, views, likes, coins, favorites, tags and url. Keep maxItems, maxCommentsPerVideo and maxDanmakuPerVideo small in agent loops – danmaku can be thousands of rows per video.

FAQ

Do I need a Bilibili account or cookies? No. The Actor uses the same public endpoints a logged-out visitor's browser uses, with anonymous visitor cookies and Bilibili's WBI request signing handled for you.

Why do I get only 20–70 comments per video? Bilibili shows logged-out visitors only the pinned and top hot comments and the first page of replies under each. The Actor collects everything a visitor can see; it does not log in, so it cannot page through all comments of a video. commentCount on the video row gives the full total.

How many danmaku can I get? The public danmaku pool of a video: up to 1,000–8,000 lines depending on video length (Bilibili's own cap). For multi-part videos the danmaku of the first part are returned.

Are commenter names exported? No. For privacy, comments and danmaku contain no author names, ids, avatars, IP regions or sender hashes – only text, time and engagement. Only public uploader (channel) names and ids are included.

I see BLOCKED error rows. Bilibili sometimes applies risk control (HTTP 412 / code -352) to datacenter IPs. The Actor already rotates visitor cookies and retries; if it persists, enable Proxy configuration → Apify Proxy → RESIDENTIAL.

Can it download videos? No. It extracts metadata, comments and danmaku only.

Does it support b23.tv short links? Not directly – open the short link once and paste the full bilibili.com/video/BV… link.

Is it legal to scrape Bilibili?

This Actor extracts only publicly available data that Bilibili shows to any logged-out visitor. It does not log in, bypass paywalls or collect personal data of commenters. Scraped content can still be copyrighted, so make sure your use complies with applicable laws (e.g. GDPR, PIPL) and Bilibili's terms of service. Do not use the data for spam or harassment. If unsure, consult a lawyer.

Related actors

Related actors by the same developer:

• Douban Scraper – Movies, Books, Ratings, Comments & Reviews
• Telegram Channel Scraper – Posts, Messages & Views, No Login
• Bluesky Scraper – Posts, Search, Profiles & Followers
• Steam Reviews & Games Scraper – Prices, Players & Top Sellers
• Google Trends Scraper – Interest Over Time & Related Queries
• Keyword Suggestions – Google, YouTube, Amazon & Bing Autocomplete

Support & feedback

Missing a field or mode? Open an issue in the Issues tab – feature requests are welcome.