Bilibili Scraper (B站 / 哔哩哔哩)

Bilibili Scraper extracts public data from Bilibili without a login, cookies or a Chinese phone number. Search videos by keyword, scrape video URLs, the trending and top-100 lists, and creator pages. Get views, likes, coins, favorites, shares, tags, top comments and danmaku (弹幕, the on-screen bullet comments) as clean JSON, CSV or Excel, or through the Apify API.

Use it for market research on Chinese audiences, influencer (UP主) discovery, trend tracking, brand monitoring, and AI/LLM pipelines that need Chinese video data.

What data can you scrape from Bilibili?

Source	Input	Output
🔍 Keyword search	searchQueries, in Chinese or English	Up to 1,000 videos per keyword, sorted by relevance, views, newest, danmaku or favorites
🎬 Video URLs	videoUrls: https://www.bilibili.com/video/BV... or a bare BV id	Full stats for each video
🔥 Trending	trending: popular or ranking	The live hot list or the all-categories top-100 leaderboard
👤 Creators (UP主)	creatorUrls: https://space.bilibili.com/<id> or the numeric id	Profile with follower counts, plus the creator's latest uploads
💬 Danmaku	includeDanmaku: true	Every bullet comment of each video, with video timestamp, colour and send time
⭐ Top comments	includeTopComments: true	The top comments of each video

Results go into three datasets, so each one is a clean table:

• Videos: the run's default dataset, one row per video.
• Creators: the dataset named creators.
• Danmaku: the dataset named danmaku.

How to scrape Bilibili

1. Click Try for free.
2. Enter search keywords, video URLs or creator URLs, or pick a trending list.
3. Turn on danmaku, top comments or full details if you need them.
4. Click Start, then download the results as JSON, CSV, Excel or HTML, or read them through the API.

Input example

{
  "searchQueries": ["原神", "iphone 17"],
  "maxResultsPerQuery": 200,
  "searchSort": "views",
  "includeDanmaku": true,
  "maxDanmakuPerVideo": 300
}

Output examples

Video

{
  "type": "video",
  "bvid": "BV12Kpxz9Eyb",
  "url": "https://www.bilibili.com/video/BV12Kpxz9Eyb",
  "title": "「小白」iPhone17全面测评：你想看的都在这了！",
  "durationSeconds": 1188,
  "publishedAt": "2025-09-17T12:00:00Z",
  "category": "数码",
  "tags": ["数码", "iPhone17", "苹果手机", "A19芯片"],
  "authorId": 8969156,
  "authorName": "小白测评",
  "views": 2106686,
  "likes": 53264,
  "coins": 27050,
  "favorites": 11884,
  "shares": 8159,
  "comments": 5844,
  "danmakuCount": 12126,
  "topComments": [{ "author": "椿巛", "text": "一时错乱了[笑哭]", "likes": 891, "replies": 8 }],
  "source": "search:iphone 17"
}

Creator

{
  "type": "user",
  "userId": 946974,
  "url": "https://space.bilibili.com/946974",
  "name": "影视飓风",
  "level": 6,
  "verifiedTitle": "2025百大UP主",
  "followers": 18638025,
  "following": 686
}

Danmaku

{
  "type": "danmaku",
  "bvid": "BV12Kpxz9Eyb",
  "text": "来了来了",
  "videoTimeSeconds": 12.5,
  "mode": "scroll",
  "color": "#ffffff",
  "sentAt": "2025-09-19T02:48:24Z"
}

How much does it cost to scrape Bilibili?

You pay only for results, with no subscription:

Event	Price	Example
Video or creator profile	$4 per 1,000 ($3.50 Silver, $3 Gold)	1,000 videos from a keyword search = $4
Danmaku line	$0.30 per 1,000 ($0.25 Silver, $0.20 Gold)	A typical video has ~180 danmaku, so about $0.05 per video
Run start	$0.00005 per run
AI danmaku digest (optional)	$0.03 per video ($0.025 Silver, $0.02 Gold)	Only when 'AI danmaku digest' is on
AI tags (optional)	$2 per 1,000 ($1.80 Silver, $1.50 Gold)	Only when 'AI tags' is on

Set a maximum cost per run and the scraper stops cleanly when it is reached. Apify's free plan includes monthly credit, enough to try about 1,000 videos.

AI insights: danmaku digest and AI tags (optional)

Turn these on when you want the meaning, not just the data. Both are off by default and cost nothing unless enabled.

• AI danmaku digest (danmakuDigest): for each video, the busiest moments with exact timestamps and comment counts, what viewers react to at each one, overall sentiment, memes and slang explained in English, and the main themes. It reads all of the video's danmaku, so you don't need to download them.
• AI tags (aiEnrichment): sentiment and score, topic labels, named entities and a one-sentence English summary for every video.

The timestamps and counts are computed exactly from the danmaku; the AI only labels them, so it can't invent numbers. Free-plan runs include a 3-item AI trial; the rest of the results are saved without AI fields.

Real output (14 October 2026, a 44-minute history video; 7,200 danmaku analysed):

{
 "type": "video",
 "bvid": "BV1Cgpd6pEBR",
 "title": "苏联版“空中监狱”是什么？【神奇组织55】",
 "danmakuDigest": {
  "summaryEn": "Viewers were highly enthusiastic and humorous, frequently using memes to react to the creator's storytelling style and the absurd historical events described. The audience engaged in a playful, mock-loyal tone, often laughing at the chaotic and ironic nature of the Soviet-era incidents presented.",
  "sentiment": "positive",
  "sentimentScore": 0.8,
  "hotMoments": [
   {
    "timeSec": 0,
    "count": 364,
    "labelEn": "Enthusiastic welcome and mock loyalty"
   },
   {
    "timeSec": 10,
    "count": 113,
    "labelEn": "Viewers celebrating the start of the video"
   },
   {
    "timeSec": 620,
    "count": 107,
    "labelEn": "Jokes about the durability of Soviet vehicle pedals"
   }
  ],
  "memes": [
   {
    "text": "忠诚",
    "meaningEn": "Loyalty; a meme used to show mock devotion to the creator."
   },
   {
    "text": "开饭",
    "meaningEn": "Dinner is served; used when a new video is uploaded."
   },
   {
    "text": "懂你意思",
    "meaningEn": "I get what you mean; used to acknowledge implied jokes or sensitive topics."
   }
  ],
  "themes": [
   "Soviet History",
   "True Crime",
   "Humor",
   "Mechanical Failures",
   "Geopolitical Irony"
  ],
  "danmakuAnalysed": 7200
 },
 "ai": {
  "sentiment": "neutral",
  "sentimentScore": 0,
  "topics": [
   "History",
   "Aviation",
   "Crime",
   "Soviet Union"
  ],
  "entities": [
   "苏联"
  ],
  "summaryEn": "This video content explores the 1990 Far East hijacking case in the Soviet Union, often referred to as the Soviet version of 'Con Air'."
 }
}

How fast is it?

• Search only: about 1,000 videos in 3 minutes.
• Full details, top comments or danmaku: about 5 seconds per video, because each extra is a separate request.

If a run gets close to its timeout, the scraper stops early and keeps everything it has already scraped. Raise the run timeout in Run options for big jobs.

Use Bilibili Scraper with AI agents (MCP)

Connect Claude, Cursor, ChatGPT or any MCP client to the Apify MCP server with this actor as a tool:

https://mcp.apify.com?tools=bossless/bilibili-scraper

The agent can then answer requests like "find the most viewed Bilibili videos about 原神 this month" or "get the danmaku of this Bilibili video". Keep agent runs small (for example maxResultsPerQuery: 20) so they finish in seconds. Videos are in the run's default dataset; creators and danmaku are in the datasets named creators and danmaku.

Use it through the API

Run the scraper and get the videos in one call:

curl -X POST "https://api.apify.com/v2/acts/bossless~bilibili-scraper/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"searchQueries": ["minecraft"], "maxResultsPerQuery": 20}'

Python:

from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("bossless/bilibili-scraper").call(run_input={"searchQueries": ["minecraft"], "maxResultsPerQuery": 20})
for video in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(video["title"], video["views"])

It also works with Make, n8n, Zapier, LangChain and other Apify integrations.

Good to know

• No login needed. The scraper uses Bilibili's public guest access through rotating residential IPs, and retries automatically when Bilibili rate-limits a request.
• Search fields. Search results don't include coins, shares or the video part id. Turn on Fetch full details to add them. Videos from URLs, creators and trending always include full stats.
• Comments. Bilibili shows logged-out visitors only the top 3 comments of a video, so topComments has at most 3 entries.
• Creator uploads. Bilibili shows logged-out visitors only a creator's most recent posts, so you get roughly their last 10 uploads. For more videos from a creator, search for their name.
• Search depth. Bilibili serves at most 1,000 results (50 pages) per keyword. Use several keywords or sort orders for wider coverage.

FAQ

Is there an official Bilibili API? Bilibili's own APIs are either internal or require a registered developer account. This scraper gives you the public data as structured JSON, without any account.

Can I scrape Bilibili comments? Yes, the top comments of every video (up to 3, because Bilibili limits logged-out visitors), plus the full danmaku stream, which is usually far richer than the comments.

Do I need proxies or cookies? No. Proxies are built in and no cookies are needed.

Can I download Bilibili danmaku (弹幕)? Yes. Turn on Include danmaku and up to maxDanmakuPerVideo bullet comments per video (default 500) are saved to the danmaku dataset, each with its video time offset, display mode, color and send time, at $0.30 per 1,000 lines (less on paid plans).

Is it legal to scrape Bilibili? This scraper collects only publicly visible data and doesn't log in. It doesn't collect the user ids of danmaku senders. Make sure your use complies with Bilibili's terms, copyright law and data-protection rules such as China's PIPL and the EU's GDPR, especially if you process personal data such as creator names.

Is this an official Bilibili product? No. This is an independent tool and is not affiliated with or endorsed by Bilibili.

Ready-made Bilibili tasks

One-click presets for common jobs. Open one, change the input if you like, and run it:

• Bilibili Danmaku AI Digest (弹幕分析) in English: hot moments, sentiment and memes explained, from all of a video's danmaku.
• Bilibili Danmaku (弹幕) Downloader: bullet comments of any video with timestamps.
• Bilibili Top 100 Ranking (排行榜) Scraper: the all-categories leaderboard, ready to schedule daily.
• Bilibili Trending & Popular Videos (热门): the live popular list with full stats.
• Bilibili Video Search by Keyword: search results with views, likes and danmaku counts.
• Bilibili Creator (UP主) Profiles & Latest Uploads: followers, verification and recent videos.

More China data scrapers

• Weibo Scraper (微博 / Sina Weibo) — Weibo hot search (微博热搜), user profiles, post search and timelines.
• Zhihu Scraper (知乎) — Zhihu hot list (知乎热榜), hot searches, column articles with full text, answers and search.
• Chinese Video Transcriber — Bilibili videos to text, English subtitles and SRT with AI.
• China Hot Search in English — Weibo, Douyin, Baidu, Zhihu and Bilibili trending topics with English summaries.

中文简介（B站爬虫 / 哔哩哔哩 API / 弹幕下载）

Bilibili Scraper（B站数据采集）：无需登录、无需 Cookie，即可采集 B站公开数据，包括关键词搜索视频、视频详情（播放、点赞、投币、收藏、分享）、标签、热门评论、弹幕、热门榜与排行榜，以及 UP主 资料和最新投稿。按结果计费，可导出 JSON、CSV、Excel，或通过 API 调用。