Bilibili Video Details Scraper

Bilibili scraper for public video data. Give it Bilibili video URLs or BV/AV IDs and get one clean row per video: title, description, uploader name and ID, publish time, duration, views, likes, coins, favourites, shares, danmaku count, comment count, tags, category, cover URL and the full part (P) list. No login, no cookies, no API key.

Use it as a lightweight Bilibili API alternative for video analytics, trend tracking, influencer research, or feeding Bilibili video data to AI agents.

Quick start

{
  "videoUrls": [
    "https://www.bilibili.com/video/BV17x411w7KC/",
    "BV1Qt411T7VS",
    "av2"
  ],
  "maxItems": 100
}

Accepted formats: https://www.bilibili.com/video/BV…, https://m.bilibili.com/video/BV…, BV17x411w7KC, av170001, 170001. Duplicates (the same video as BV and AV) are removed. b23.tv short links are not resolved; open them and paste the full URL.

Output

One dataset row per video (JSON, CSV, Excel via Apify datasets):

Field	Meaning
bvid, aid, url	BV ID, numeric AV ID, canonical video URL
requested	The input entry that produced the row
title, description	Video title and description
uploaderId, uploaderName	Uploader's public user ID (mid) and display name
publishedAt, publishTimestamp	Publish time (ISO 8601 UTC / Unix seconds)
durationSeconds	Total duration
views, likes, coins, favorites, shares	Engagement counters
danmakuCount, commentCount	Bullet-comment and comment counts (texts not collected)
tags	Public tag names (background-music tags excluded)
categoryId, categoryName, parentCategoryName, legacyCategoryId	Category (zone)
coverUrl	Cover image (HTTPS)
isOriginal	Original upload (true) or repost (false)
partCount, parts	Number of parts and [{page, cid, title, durationSeconds}]
fetchedAt	Fetch time

Example row (abridged):

{"bvid":"BV17x411w7KC","aid":170001,"title":"【MV】保加利亚妖王AZIS视频合辑","uploaderId":122541,"uploaderName":"冰封.虾子","publishedAt":"2011-11-09T14:55:33.000Z","durationSeconds":2412,"views":45840786,"likes":943698,"coins":295112,"favorites":876828,"shares":44354991,"danmakuCount":915692,"commentCount":184255,"tags":["保加利亚妖王","保加利亚","Азис","azis","mv"],"categoryId":2017,"categoryName":"MV","parentCategoryName":"音乐","partCount":10}

The SUMMARY record in the run's key-value store lists totals, the stop reason, every video that failed (removed, private, region-restricted or invalid input) and network bytes.

Pricing (proposed, not active)

Proposed pay-per-event price: $3.00 per 1,000 videos ($0.003 per result). You would pay only for videos returned in the dataset; invalid IDs, removed or region-restricted videos and failed requests are never charged. The run stops cleanly when your spending limit is reached. This pricing is a proposal and is not configured on the Actor yet.

How it works

• Reads the public, server-rendered video page www.bilibili.com/video/<BV>/ (allowed by Bilibili's robots.txt) and stops downloading as soon as the embedded video data is complete.
• Bilibili answers HTTP 412 to most cloud IPs. The Actor tries a direct request first and switches to the proxy from proxyConfiguration (default: Apify residential, China) only after a 412.
• Polite by default: 2 parallel requests, at least 400 ms between requests, exponential backoff and session rotation on errors.

Limits

• Video details only. Popular, ranking and search lists are not included: they are served only by api.bilibili.com, whose robots.txt disallows all automated access. Collect IDs another way (e.g. from the website) and pass them in.
• No login, cookies, CAPTCHA solving, comment texts, danmaku texts, follower counts or uploader profiles. Members-only, paid, private and region-locked videos are reported as failed in SUMMARY.
• Counters are a snapshot at fetch time. categoryName comes from the page's own category list and can be null for retired categories.
• You are responsible for how you use and redistribute the data under Bilibili's terms and applicable law.