Bilibili Scraper — Trending Videos, Views & Engagement

China's largest video community in rows: title and description, uploader and channel ID, publish date, duration, and the full engagement set — views, likes, coins, favourites, comments, shares and danmaku.

What does the Bilibili Scraper do?

This Actor collects videos from Bilibili, the video community at the centre of Chinese youth internet culture — anime, gaming, technology, music and study content, with a user base that behaves nothing like a western platform's.

That difference is why this Actor collects the metrics it does. Bilibili counts coins: a limited daily token a viewer spends on a video they think deserves it. Spending one costs the viewer something, so the coin count is a far better signal of how well a video actually landed than views. It also counts danmaku — the comments that scroll across the frame — which measures live engagement rather than after-the-fact commenting. Both are collected alongside the usual views, likes, favourites, comments and shares.

Who is it for?

• Social media analysts tracking what trends in the Chinese market.
• Brand and agency teams identifying creators to work with on Bilibili.
• Market researchers studying Chinese consumer and youth culture.
• Games and anime publishers watching reception of a title in China.
• Creator-economy platforms benchmarking channel performance.
• Academics studying engagement on a non-western social platform.

Use cases

• Track what is trending on Bilibili day by day.
• Rank videos by coins rather than views to find genuinely well-received content.
• Find high-performing creators in a category for outreach.
• Compare danmaku volume against comment count to measure live engagement.
• Measure how quickly a video accumulates engagement after publishing.
• Build a dataset of Chinese video titles and descriptions for text analysis.
• Feed an AI agent trending Chinese video data for market questions.

Why use this Bilibili Scraper?

• The full engagement set — views, likes, coins, favourites, comments, shares and danmaku — not just views.
• Uploader name and channel ID, so creators can be tracked across runs.
• Publish dates converted to ISO from Bilibili's unix timestamps.
• Duration in seconds, ready for engagement-per-minute analysis.
• Fast and proxy-free — the whole popular feed in about ten seconds.
• Keyless and login-free.

What data can you extract?

One row per video. Anything the source left blank comes back as null.

Field	Description
bvid	Bilibili video ID
url	Link to the video
title	Video title
description	Video description
uploader	Uploader display name
uploaderId	Uploader channel ID (mid)
publishedAt	Publish date as ISO
durationSec	Duration in seconds
views	View count
likes	Like count
coins	Coins spent on the video
favourites	Times saved to a favourites list
comments	Comment count
shares	Share count
danmaku	Scrolling on-screen comments
category	Bilibili category name
thumbnail	Thumbnail image
scrapedAt	ISO timestamp of collection

Sample output

{
  "bvid": "BV1G48M6XEBt",
  "url": "https://www.bilibili.com/video/BV1G48M6XEBt",
  "title": "\"复活吧！我的（ ）\"",
  "description": "致准星上的微光，重逢的微风和那个夏天。 2026.8.21",
  "uploader": "HL-Flame",
  "uploaderId": "3546883796503085",
  "publishedAt": "2026-08-21T04:30:45.000Z",
  "durationSec": 250,
  "views": 2799365,
  "likes": 350110,
  "coins": 221412,
  "favourites": 177810,
  "comments": 12190,
  "shares": 27232,
  "danmaku": 2482,
  "category": "网络游戏",
  "thumbnail": "http://i0.hdslb.com/bfs/archive/32bd89a388d516e86c9aef366921.jpg",
  "scrapedAt": "2026-08-22T08:30:28.040Z"
}

How to use the Bilibili Scraper

Option A — the popular feed

Leave Feed on Popular, set Maximum results and run. The feed is a rolling window of several hundred videos and a run collects the lot in seconds.

Option B — the top chart

Set Feed to Top ranking chart for Bilibili's fixed ranking. It is a single chart rather than a paged feed, so it returns a set number of videos.

Option C — a daily trend series

Schedule a daily run on the popular feed and keep the results; diffing by bvid shows what entered and left the feed.

Input parameters

Parameter	Type	Default	Description
feed	select	popular	popular (paged) or ranking (fixed chart)
maxResults	integer	1000	Stop after this many videos (max 20,000)
proxyConfiguration	object	(off)	Optional; the public API answers without one

Tips for best results

• Rank by coins, not views. A coin costs the viewer a limited daily token, so it separates genuine enthusiasm from passive watching.
• Normalise by duration. views / durationSec and danmaku / durationSec compare a three-minute clip to a half-hour video fairly.
• uploaderId is the stable creator key; display names change, the mid does not.
• The popular feed is a window, not an archive. It holds a few hundred videos at any moment, so schedule daily runs if you want a trend series rather than a single snapshot.
• Turn on a proxy only if you see rate limiting — the API normally answers fine without one.

Integrations

Connect the dataset to Make, Zapier, Airbyte, Google Sheets, Slack, GitHub or any HTTP endpoint through Apify integrations, or schedule a run and push results into a warehouse with a webhook.

API usage

from apify_client import ApifyClient

client = ApifyClient("<YOUR_API_TOKEN>")
run = client.actor("logiover/bilibili-video-scraper").call(run_input={
    "feed": "popular",
    "maxResults": 2000,
})
rows = list(client.dataset(run["defaultDatasetId"]).iterate_items())
best = sorted(rows, key=lambda r: (r["coins"] or 0) / max(r["views"] or 1, 1), reverse=True)
print("highest coin rate:", best[0]["title"], best[0]["uploader"])

Use with AI agents (MCP)

The Actor is callable from the Apify MCP server, so an assistant can pull trending Bilibili videos directly and answer questions about Chinese video trends and creators.

FAQ

Do I need a Bilibili account or API key?

No. This Actor reads Bilibili's public web interface; there is nothing to register.

How many videos can one run return?

The popular feed is a rolling window rather than an archive: it holds roughly 600-700 videos at a time, twenty per page, and a run collects all of them in about ten seconds. Setting a higher maximum does no harm — the run simply stops when the feed is exhausted. The ranking chart is a fixed list and returns however many videos Bilibili puts on it.

What are coins, and why do they matter?

A coin is a limited daily token a Bilibili user can spend on a video they rate highly. Because spending one has a real cost to the viewer, the coin count is the platform's strongest quality signal — much stronger than views, which cost nothing.

What is danmaku?

Comments that scroll across the video as it plays, rather than sitting underneath it. The count measures engagement during viewing, which is a different thing from the comment count.

Are titles and descriptions in Chinese?

Yes. Bilibili is a Chinese-language platform, so titles, descriptions and category names come back in Chinese, exactly as published.

Can I scrape a specific creator's channel?

Not with this Actor. It collects the trending and ranking feeds, which is where discovery happens. Every row carries the uploaderId, so creators found here can be tracked across runs.

Do I need a proxy?

Usually not — the public API answers directly. Enable one if you are running very large exports and start seeing rate limiting.

Can I export to CSV or Excel?

Yes — every run's dataset exports to JSON, CSV, Excel, XML or JSONL.

How fresh is the data?

Each run reads the feed live, so engagement counts are as current as Bilibili's own at the moment you run it.

Why did my run return zero videos?

Usually rate limiting from the exit you are running on. Enable a proxy and try again; the Actor logs Bilibili's own error code when that happens.

Is it legal to scrape Bilibili?

This Actor reads Bilibili's public web interface — no accounts, no logins, no private content. Video titles, engagement counts and public channel names are information published for public view. Uploader names are public identities but may still be personal data, so you remain responsible for how you use the results, including under the GDPR, and for respecting the platform's terms in your own jurisdiction.

Related scrapers

• YouTube Scraper suite — video, channel and transcript data.
• TikTok Scraper suite — video and hashtag data.
• Threads Scraper — posts and engagement.
• Bluesky Historical Archive Scraper — full-feed social archives.