YouTube Channel Scraper Pro 🚀

A high-performance, reliable Apify Actor for scraping YouTube channel videos — including Long-Form and Shorts — with advanced filtering and infinite scroll support.

🔗 Apify Actor URL: https://apify.com/skcho/youtube-channel-scraper-pro

---

✨ Key Features

• 🔄 Infinite Scroll (Pagination): Automatically scrolls through the entire channel page using a real browser — no 30-video limit.
• 🗂️ Long-Form & Shorts Separation: Precisely scrape and categorize videos by type.
• 📦 Rich Data Extraction: Titles, View Counts, Upload Dates, Durations, HQ Thumbnails, and Animated WebP Previews — all extracted out of the box.
• 👍 Optional Like Count Extraction: A lightweight, high-concurrency HTTP-based fetcher grabs like counts at ~430 videos/min without launching extra browsers.
• 🎯 Precision Filtering: Filter by minimum views, likes, upload date range, and more.
• 🧠 Hybrid Architecture: PlaywrightCrawler handles real-browser scrolling; HttpCrawler handles likes — minimizing memory usage and maximizing speed.

---

⚡ Performance

Scenario	Speed
Channel scroll + data extract (100 videos)	~15 seconds
Like count fetch (per 100 videos, fetchLikes: true)	~15 seconds
Total for 100 videos with likes	~30 seconds

Likes are fetched at ~430 requests/min in parallel via lightweight HTTP requests — no extra browser needed.

---

🛠️ Input Parameters

Parameter	Type	Default	Description
channelUrls	Array<String>	(Required)	YouTube channel URLs, channel IDs, or @handles to scrape (e.g., https://www.youtube.com/@mkbhd, @veritasium). Multiple channels supported.
collectChannelInfo	Boolean	true	Collect channel profile metadata (name, description, subscribers, total views, video count, joined date, country, avatar, banner, links).
includeVideos	Boolean	false	Also collect video rows from the channel's videos/shorts tabs. Leave disabled when you only need channel profile info.
videoType	Enum	"ALL"	Which video type to scrape: "ALL", "LONG_FORM", or "SHORTS".
minUploadDate	String	""	Only include videos uploaded on or after this date (approximate — derived from YouTube's relative date text). Format: YYYY-MM-DD.
maxUploadDate	String	""	Only include videos uploaded on or before this date (approximate). Format: YYYY-MM-DD.
maxDaysOld	Integer	0	Stop scraping once videos are older than this many days. 0 disables the check.
minViews	Integer	0	Only include videos with at least this many views. (e.g., 1000000 for 1M+).
maxItemsPerChannel	Integer	5	Maximum videos to scrape per channel (per tab). 0 means unlimited. Uses infinite scroll to exceed YouTube's default page limit.
fetchLikes	Boolean	false	If enabled, fetches exact Like counts for every video via fast parallel HTTP requests. Adds ~15s per 100 videos.
minLikes	Integer	0	(Requires fetchLikes: true) Only include videos with at least this many likes.
proxyUrls	Array<String>	[]	Optional list of proxy URLs (e.g. http://user:pass@host:port) to rotate through. Recommended for anything beyond a handful of channels — without a proxy, YouTube can rate-limit or temporarily block your IP. Leave empty to run without a proxy.
collectComments	Boolean	false	(Requires includeVideos: true) Collects comments (and optionally replies) for each collected video as separate recordType: "comment" records.
maxCommentsPerVideo	Integer	50	Maximum comments (including replies, if enabled) to collect per video.
includeCommentReplies	Boolean	true	Also collect replies to each top-level comment.
commentSortBy	Enum	"TOP_COMMENTS"	"TOP_COMMENTS" or "NEWEST_FIRST".

---

📄 Output Data Structure

Each video is pushed as a JSON object to the Apify Dataset:

{
  "channelName": "Marques Brownlee",
  "channelUrl": "https://www.youtube.com/@mkbhd",
  "videoId": "eFeDpUVEy48",
  "videoUrl": "https://www.youtube.com/watch?v=eFeDpUVEy48",
  "title": "The Biggest Android Update Ever",
  "type": "LONG_FORM",
  "uploadDateText": "8 days ago",
  "viewCount": 4099999,
  "durationText": "12:59",
  "thumbnailUrl": "https://i.ytimg.com/vi/eFeDpUVEy48/hqdefault.jpg",
  "animatedThumbnailUrl": "https://i.ytimg.com/an_webp/eFeDpUVEy48/mqdefault_6s.webp",
  "likeCount": 125000
}

Field Reference

Field	Type	Description
channelName	String	Official channel display name
channelUrl	String	Input channel URL
videoId	String	YouTube video ID
videoUrl	String	Full youtube.com/watch?v=... URL
title	String	Video title
type	String	"LONG_FORM" or "SHORTS"
uploadDateText	String	Relative upload date (e.g., "8 days ago")
viewCount	Integer	Parsed integer view count
durationText	String	Video duration string (e.g., "12:59")
thumbnailUrl	String	Highest-quality static thumbnail URL
animatedThumbnailUrl	String	3-second animated WebP preview URL
likeCount	Integer	Like count — present only when fetchLikes: true

Need subtitles/captions? That's now a separate module — see youtube-subtitle-downloader/.

Comment records (recordType: "comment", when collectComments: true)

Field	Type	Description
videoId / videoUrl	String	Which video this comment belongs to
commentId	String	YouTube comment ID
parentCommentId	String	Present only on replies
isReply	Boolean	true for replies
authorName / authorChannelId	String	Commenter's display name and channel ID
authorIsChannelOwner	Boolean	true if the video's own channel posted this comment
text	String	Comment text
likeCount / replyCount	Integer	Parsed counts
isPinned	Boolean	Pinned-comment flag

---

⚙️ How It Works (Architecture)

This actor uses a two-phase hybrid architecture to maximize both stability and speed:

Phase 1 — Scroll & Extract (PlaywrightCrawler)

A real Chromium browser navigates to the channel's /videos and /shorts tabs. It automatically scrolls down the page, triggering YouTube's infinite scroll to load more videos, until maxItemsPerChannel is reached or the page end is detected. Video metadata is then extracted directly from the rendered DOM.

Browser optimizations:

• Blocks unnecessary resources (images, fonts, media) to reduce memory usage
• Disables GPU, extensions, and background services
• Anti-bot fingerprint masking (navigator.webdriver suppressed)

Phase 2 — Like Count Fetch (HttpCrawler, Optional)

If fetchLikes is enabled, up to 20 concurrent lightweight HTTP requests are made to individual youtube.com/watch?v=... pages. Like counts are extracted by parsing the embedded ytInitialData JSON object — no browser required. This runs at ~430 requests/min.

---

🤝 Support & Issues

Found a bug or have a feature request? Please reach out via the Apify Issues tab!

cd C:\Users\skps\Documents\GitHub\youtube-channel-scraper-bundle\gui-platform

npm install node server.js