Bilibili Scraper - Videos, Search, Popular & Uploader Stats

Bilibili Scraper pulls public data from Bilibili (bilibili.com, B站), China's biggest video community. It reads the popular (热门) feed, runs keyword searches and looks up specific videos. Each video comes back with views, likes, coins, favorites, shares, danmaku and comment counts, plus its uploader. You can also get uploader (UP主) follower stats and, if you want them, each video's top comments, all from one Actor. It costs $0.002 per video, 10x less than the most-used Bilibili Actor ($0.02 per item). An onlyNew monitoring mode outputs (and charges) only videos you haven't seen before. If Bilibili holds a field back, the item says so in a status field instead of leaving it blank without explanation. No login, no cookies from an account, no browser.

What data can you get?

Videos (type: "video")	Uploaders (type: "uploader")	Comments (type: "comment", optional)
BV id, AV id (aid), video URL	mid (user id), space URL	Comment id and link
Title, description, cover image	Name, bio (sign), avatar	Comment text
Duration (seconds), publish date (ISO)	Followers (fans), following	Likes, reply count
Category (zone / tname), tags (search)	Number of videos uploaded	Time (ISO)
Uploader mid, name and space URL	Total likes received	Commenter display name + mid (public handle)
Views, danmaku, likes, coins, favorites, shares, comments	Account level (0-6)	Video BV id and aid
Position in the popular feed / search results	Official verification title
detailsStatus (full / partial), commentsStatus

Only public creator and video data is collected. The Actor collects no e-mail addresses, phone numbers, gender, birthday or location.

Use cases

• Trend monitoring: save the Bilibili popular (热门) feed every day and see which topics, games and creators are taking off in China.
• Brand and product research: search your brand, game or product (原神, iPhone 17, Tesla) and measure views, likes and coins on the videos about it.
• Influencer (KOL) discovery and vetting: compare UP主 follower counts, video counts, total likes and verification before a campaign.
• Competitor tracking: schedule a search sorted by newest with onlyNew to get only new videos about a keyword each day.
• Content and sentiment research: collect top comments for a quick read of the audience reaction.

How to use it

1. Pick a Mode: popular, search, videos or uploaders.
2. Fill in the matching field: Search keywords, Video URLs or BV ids, or Uploader mids or space URLs.
3. Set Max items (default 50; 0 = no limit).
4. Optional: turn on Include top comments, or Only new videos for scheduled monitoring.
5. Click Start, then download JSON, CSV or Excel, or read the dataset through the API.

Popular feed (the daily health-check input):

{ "mode": "popular", "maxItems": 10 }

Search videos by keyword, newest first, only new ones on each scheduled run:

{ "mode": "search", "keywords": ["原神", "minecraft"], "searchOrder": "pubdate", "maxItems": 50, "onlyNew": true, "stateKey": "games-monitor" }

Specific videos with their top comments:

{ "mode": "videos", "videoUrls": ["https://www.bilibili.com/video/BV1Ev4w63EW4/", "BV1zUx16tEKg"], "includeComments": true }

Uploader (UP主) stats:

{ "mode": "uploaders", "uploaders": ["546195", "https://space.bilibili.com/927587", "517327498"] }

Output example

Real items from a local run on 3 Nov 2026 (one video, one uploader and one comment; the commenter's name is replaced here):

[
  {
    "type": "video",
    "id": "BV1zUx16tEKg",
    "aid": 117535236296957,
    "url": "https://www.bilibili.com/video/BV1zUx16tEKg/",
    "title": "180万一平米！假装要给男友买义乌1700万9平米的爆火房…他震惊了！",
    "description": "180万一平米！！！震惊了！！！",
    "coverUrl": "https://i1.hdslb.com/bfs/archive/c3938ea496ad80f58972e8e394fcf2e53ada39ce.jpg",
    "durationSec": 613,
    "publishedAt": "2026-11-01T03:00:00.000Z",
    "category": "日常",
    "uploaderMid": "2342793",
    "uploaderName": "是王小白呀",
    "uploaderUrl": "https://space.bilibili.com/2342793",
    "views": 555971,
    "danmaku": 1660,
    "likes": 14755,
    "coins": 2167,
    "favorites": 1993,
    "shares": 1207,
    "replies": 372,
    "tags": null,
    "source": "popular",
    "searchKeyword": null,
    "position": 1,
    "detailsStatus": "full",
    "commentsStatus": "ok",
    "scrapedAt": "2026-11-03T10:01:16.895Z"
  },
  {
    "type": "uploader",
    "id": "546195",
    "url": "https://space.bilibili.com/546195",
    "name": "老番茄",
    "sign": "新浪微博：_老番茄_",
    "fans": 20867223,
    "following": 7,
    "archiveCount": 683,
    "likes": 196657353,
    "level": 6,
    "officialVerify": "2025百大UP主、2025年度商业影响力奖UP主、2022年度联合创作奖UP主",
    "avatarUrl": "https://i0.hdslb.com/bfs/face/bc5ca101313d4db223c395d64779e76eb3482d60.jpg",
    "scrapedAt": "2026-11-03T10:01:20.699Z"
  },
  {
    "type": "comment",
    "id": "319067527793",
    "url": "https://www.bilibili.com/video/BV1zUx16tEKg/#reply319067527793",
    "videoId": "BV1zUx16tEKg",
    "aid": 117535236296957,
    "text": "500万涨到1700万，这是在洗钱还是在干什么见不得人的…",
    "likes": 35,
    "replyCount": 10,
    "publishedAt": "2026-11-03T08:06:21.000Z",
    "commenterName": "commenter1",
    "commenterMid": null,
    "scrapedAt": "2026-11-03T10:01:17.393Z"
  }
]

The dataset has three views: Videos, Uploaders and Comments. A RUN_SUMMARY record in the key-value store lists each input's status, the item counts, the request count and any warnings.

Pricing

Pay per result. You pay only for items saved to the dataset:

Event	Price
video - one video	$0.002
uploader - one uploader profile	$0.002
comment - one comment	$0.0003

Worked examples: 1,000 videos ≈ $2.00. 100 popular videos with their top 3 comments = 100 × $0.002 + 300 × $0.0003 = $0.29. 50 uploader profiles = $0.10. In onlyNew mode, videos you have already seen are skipped and not charged. The run stops cleanly when it reaches your Maximum cost per run.

Monitoring new videos (onlyNew)

Turn on Only new videos in popular or search mode and schedule the Actor (daily, hourly). The first run outputs everything and remembers it. Later runs with the same Monitor name (state key) output only videos not seen before. For keyword monitoring, use "searchOrder": "pubdate" (newest first). Use a different state key for each independent monitor.

Tips and limits

• Comments: at most ~3 per video. Bilibili shows visitors who are not logged in only the top ~3 hot comments per video, and this Actor never logs in. maxCommentsPerVideo is an upper limit. Videos where Bilibili returned fewer comments than you asked for get commentsStatus: "limited". Each video with comments costs one extra request.
• Search results have no coins or shares. Bilibili's search API does not return them, so search items have coins/shares = null and detailsStatus: "partial". Popular-feed and video-mode items have all stats ("full").
• Tags are only filled for search results (the popular and video-detail endpoints don't include them); elsewhere tags is null.
• Video mode on cloud IPs: Bilibili's video-detail endpoint sometimes answers HTTP 412 (risk control) to datacenter IPs. The Actor then looks each video up through search instead: those items have source: "search-fallback" and detailsStatus: "partial", and the run summary includes a warning. AV-only ids can't use the fallback, so pass BV ids.
• Limits: the popular feed has a few hundred videos. Search returns up to 50 pages (about 1,000 videos) per keyword. maxItems applies per keyword in search mode.
• Speed and politeness: at most 2 requests per second. About 20 videos per request in popular and search mode.
• Deleted, private or region-locked videos are reported as failed inputs. The rest of the run carries on.

FAQ

Is it legal to scrape Bilibili? This Actor reads only public data that anyone can see without logging in: video stats, public creator profiles, and the top comments with their public display names. It does not log in, solve captchas or bypass access controls, and it collects no contact details. Bilibili's API robots.txt disallows crawling, so keep volumes moderate and use the data responsibly. You are responsible for complying with Bilibili's terms, copyright and data-protection law (such as GDPR or China's PIPL) in your use case. If in doubt, ask a lawyer.

Do I need a Bilibili account or cookies? No. The Actor opens the Bilibili homepage once to get the anonymous visitor cookie (buvid3) that every first-time visitor receives, then calls the public JSON APIs.

Do I need a proxy? Usually not: the popular feed and uploader cards work from Apify's datacenter IPs. If a run reports "Bilibili risk control refused the request (HTTP 412)", enable Apify Proxy (datacenter). Residential proxies are not supported: if you select them, the Actor falls back to Apify datacenter proxy and logs a warning.

Why are some fields null? Bilibili doesn't return them on that endpoint (see Tips). detailsStatus and commentsStatus tell you which case you're in, so a missing value is never silent.

Can I get danmaku (bullet comments) text, subtitles or the video file? No. The Actor reports the danmaku count and video metadata, not video files or danmaku text.

Integrations

• API: start runs and read datasets with the Apify API or the JavaScript/Python clients.
• Make, Zapier, n8n: use the Apify modules to start the Actor on a schedule and send new videos to Google Sheets, Slack or a database.
• MCP / AI agents: call the Actor from Claude, ChatGPT or any MCP client through Apify's MCP server.
• Webhooks: get notified when a run finishes.

Changelog

• 0.1 (Nov 2026): first release with popular, search, videos and uploaders modes, optional top comments, onlyNew monitoring and per-event pricing.

More scrapers by the same developer

• AI Visibility Tracker: ChatGPT, Perplexity, Gemini, Claude
• Google Ads Transparency Scraper
• Substack Scraper
• Telegram Channel Scraper
• ATS Jobs Scraper: Greenhouse, Lever, Ashby & Workday
• Dice Jobs Scraper

Browse all my Actors on my Apify Store profile.

Support

Found a bug or need a field? Open an issue on the Actor's Issues tab. Include the run ID and your input.