YouTube Scraper Pro: Videos, Channels, Comments & Transcripts

Scrape YouTube videos, channels, playlists, shorts, and live streams. Each row ships video metadata, channel intelligence, engagement metrics, transcripts in multiple languages, top comments with replies, chapters, most replayed heatmap, related videos, and music credits. 16 regions, 16 languages, no API quota. Pay per video.

Built for content marketers tracking competitor performance, brand teams monitoring mentions across creators, AI researchers harvesting transcripts for training data, journalists investigating viral content, sponsorship platforms benchmarking creator audiences, social listening tools, and BI teams piping YouTube catalog data into a warehouse.

Keywords this actor ranks for: youtube scraper, youtube api alternative, youtube data extractor, youtube video scraper, youtube channel scraper, youtube comments scraper, youtube transcript scraper, youtube shorts scraper, youtube playlist scraper, youtube hashtag scraper, youtube to JSON, youtube to CSV, youtube most replayed, youtube heatmap api, youtube chapters api.

---

Why this actor

Other YouTube scrapers	This actor
Search terms only	Five input modes: search terms, URLs, video IDs, channel handles, playlists
Title and view count only	Full engagement: likes, comments, engagement rate, view count
English transcripts only	Multi language transcripts with priority fallback
No comment data	Top comments with author handle, like count, pinned + hearted flags, replies
No heatmap	Most replayed segments (timestamp + intensity 0 to 1)
No chapters	Chapter timeline with start times and thumbnails
Single region	16 regions, 16 interface languages
No channel enrichment	Subscriber count, video count, country, links, keywords
Skips music credits	Song title, artist, album, license from the music panel
Search filters lacking	Sort by recent / view count / rating, filter by 4K / live / HD / CC license

---

How it works

flowchart LR
    A[Search terms<br/>URLs<br/>Video IDs<br/>Channel handles<br/>Playlists] --> B[Listing crawl<br/>collect video IDs]
    A --> C[Direct watch<br/>per video ID]
    B --> C
    C --> D[Watch page<br/>Playwright + fingerprinted Chrome]
    D --> E[Parse ytInitialData<br/>+ ytInitialPlayerResponse]
    E --> F[Enrichment fetches<br/>transcript + comments + channel stats]
    F --> G[One row per video]
    G --> H[(JSON CSV API)]

Pages render with Playwright behind rotating fingerprints. YouTube consent walls dismiss automatically. Channel stats resolve once per channel and cache across the run.

---

What you get per row

flowchart LR
    R[Video row] --> R1[Identity<br/>videoId url type duration]
    R --> R2[Channel<br/>name handle subs verified]
    R --> R3[Engagement<br/>views likes comments rate]
    R --> R4[Transcript<br/>multi language timed segments]
    R --> R5[Comments<br/>top thread replies pinned]
    R --> R6[Chapters<br/>timestamps and thumbnails]
    R --> R7[Heatmap<br/>most replayed intensity]
    R --> R8[Music credits<br/>song artist album]
    R --> R9[Related videos<br/>up next sidebar]
    R --> R10[End screens<br/>cards info cards]

Toggle on extractStreamUrls and the row carries the available video and audio stream URLs and bitrates.

---

Quick start

Scrape a search with 4K filter, sorted by view count

{
  "searchTerms": ["drone footage iceland"],
  "maxVideosPerSearch": 50,
  "sortBy": "viewCount",
  "features": ["fourK"],
  "uploadDate": "year"
}

Direct video IDs with transcripts and chapters

{
  "videoIds": ["arj7oStGLkU", "ZbZSe6N_BXs", "AKJfakEsgy0"],
  "extractTranscript": true,
  "transcriptLanguages": ["en", "es"],
  "extractChapters": true,
  "extractMostReplayed": true
}

Track a creator's catalog with date range

{
  "channelHandles": ["@MrBeast"],
  "maxVideosPerChannel": 50,
  "dateFrom": "2026-01-01",
  "extractTopComments": true,
  "maxComments": 100
}

Multi region listening (US + Brazil + Japan)

{
  "searchTerms": ["world cup highlights"],
  "region": "BR",
  "language": "pt",
  "maxVideosPerSearch": 30
}

Playlist export with engagement and channel stats

{
  "startUrls": ["https://www.youtube.com/playlist?list=PLrAXtmRdnEQy6nuLMHjMZOz59Oq8B9oT9"],
  "maxVideosPerPlaylist": 200,
  "extractChannelStats": true,
  "extractRelatedVideos": true
}

---

Sample output

{
  "videoId": "arj7oStGLkU",
  "url": "https://www.youtube.com/watch?v=arj7oStGLkU",
  "embedUrl": "https://www.youtube.com/embed/arj7oStGLkU",
  "type": "video",
  "title": "Inside the Mind of a Master Procrastinator | Tim Urban | TED",
  "description": "Tim Urban knows that procrastination doesn't make sense...",
  "durationSeconds": 836,
  "durationText": "13:56",
  "publishDate": "2016-04-06T00:00:00-07:00",
  "category": "Education",
  "keywords": ["TED", "TED Talk", "Tim Urban", "procrastination"],
  "channel": {
    "name": "TED",
    "channelId": "UCAuUUnT6oDeKwE6v1NGQxug",
    "handle": "@TED",
    "url": "https://www.youtube.com/@TED",
    "subscriberCount": 25700000,
    "videoCount": 4521,
    "isVerified": true,
    "country": "US"
  },
  "engagement": {
    "viewCount": 61194372,
    "likeCount": 1372411,
    "commentCount": 32198,
    "engagementRate": 0.02243
  },
  "flags": {
    "isLive": false,
    "isShort": false,
    "isFamilySafe": true,
    "hasCaptions": true,
    "hasChapters": true,
    "hasMostReplayed": true
  },
  "thumbnails": {
    "default": "https://i.ytimg.com/vi/arj7oStGLkU/default.jpg",
    "high": "https://i.ytimg.com/vi/arj7oStGLkU/hqdefault.jpg",
    "maxres": "https://i.ytimg.com/vi/arj7oStGLkU/maxresdefault.jpg"
  },
  "chapters": [
    { "title": "The procrastinator's brain", "startSeconds": 142, "thumbnail": "https://..." },
    { "title": "The Panic Monster", "startSeconds": 412, "thumbnail": "https://..." }
  ],
  "mostReplayed": [
    { "startSeconds": 388.4, "durationSeconds": 8.36, "intensityScore": 1.0 },
    { "startSeconds": 412.7, "durationSeconds": 8.36, "intensityScore": 0.92 }
  ],
  "transcript": {
    "language": "en",
    "isAutogenerated": false,
    "segments": [
      { "start": 0.0, "duration": 4.2, "text": "So in college I was a government major..." }
    ],
    "text": "So in college I was a government major, which means..."
  },
  "topComments": [
    {
      "author": "@viewer123",
      "channelHandle": "@viewer123",
      "text": "This talk changed my life",
      "likeCount": "12K",
      "pinned": false,
      "hearted": true,
      "replyCountText": "48 replies"
    }
  ],
  "scrapedAt": "2026-04-26T08:00:00.000Z"
}

---

Who uses this

Role	Use case
Content marketer	Audit a competitor's channel. Pull every video with views, likes, engagement rate, and chapter topics.
Brand monitor	Search terms across mentions of your brand. Watch for sentiment shifts via comments.
AI / RAG team	Pull transcripts from a topic cluster. Each segment has timecodes for citation.
Sponsorship platform	Benchmark creator audiences. Subscriber count, country, video frequency, engagement rate.
News and research	Investigate viral patterns. Most replayed heatmap shows where viewers rewind.
Music labels	Track which shorts use a song. Music credits panel returns title, artist, album, license.
Educator / publisher	Export playlists with chapter timestamps for course mapping.
BI analyst	Pipe YouTube catalog data into Snowflake. Each row is API ready.

---

Input reference

Field	Type	What it does
searchTerms	string[]	YouTube search queries.
startUrls	string[]	Mix watch, channel, playlist, shorts, hashtag, results URLs.
videoIds	string[]	Direct 11 character video IDs. Fastest input mode.
channelHandles	string[]	Channel handles like @MrBeast.
maxVideosPerSearch	integer	Cap on regular videos per search.
maxShortsPerSearch	integer	Cap on shorts per search.
maxStreamsPerSearch	integer	Cap on live streams per search.
maxVideosPerChannel	integer	Cap when scraping channel catalogs.
maxVideosPerPlaylist	integer	Cap when scraping playlists.
uploadDate	enum	hour, today, week, month, year, any.
duration	enum	short (<4m), medium (4-20m), long (>20m), any.
sortBy	enum	relevance, uploadDate, viewCount, rating.
features	string[]	live, fourK, hd, subtitles, creativeCommons, vr360, hdr.
dateFrom, dateTo	string	ISO date bounds for channel and playlist scraping.
extractTranscript	boolean	Multi language transcript with timecodes.
transcriptLanguages	string[]	Priority list of language codes.
extractTopComments	boolean	Top comment thread with author and like count.
maxComments	integer	Cap on top level comments per video.
extractCommentReplies	boolean	Walk into each thread and pull replies.
extractChapters	boolean	Chapter timestamps and thumbnails.
extractRelatedVideos	boolean	Up next sidebar suggestions.
extractChannelStats	boolean	Subscriber count, video count, country, links.
extractMusicAndCredits	boolean	Song title, artist, album, license.
extractMostReplayed	boolean	Heatmap segments with intensity scores.
extractStreamUrls	boolean	Direct video and audio stream URLs.
extractEndScreens	boolean	End screen elements and info cards.
region	enum	gl param. 16 regions including US, GB, JP, BR, IN, KR, DE.
language	enum	hl param. 16 interface languages.
dedupe	boolean	Skip video IDs already pushed across runs.
concurrency	integer	Parallel pages. Three to five is safe.
proxyConfiguration	object	Apify proxy. Datacenter is fine. Residential for high volume.

---

API call

curl -X POST \
  "https://api.apify.com/v2/acts/YOUR_USER~youtube-scraper/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "channelHandles": ["@MrBeast"],
    "maxVideosPerChannel": 20,
    "extractTranscript": true,
    "extractMostReplayed": true,
    "extractChapters": true
  }'

---

Pricing

The first few videos per run are free so you can validate the output before paying. After that, one charge per video row. Chapters, channel stats, related videos, end screens, and most replayed are included at no extra cost. Transcripts and comments add modest extra processing.

---

FAQ

What is the difference between this and the YouTube Data API?

YouTube's official API has a 10,000 unit per day quota. One search costs 100 units, one comment fetch costs 1 unit. Realistically you get a few thousand videos per day. This actor has no quota and no API key. Plus you get fields the official API does not expose: most replayed heatmap, full transcripts, music credits, chapters with thumbnails, and end screen cards.

Does it pull transcripts in non English languages?

Yes. Pass transcriptLanguages: ["es", "en"] to prefer Spanish and fall back to English. The actor walks the available caption tracks for the video and returns the first match. Auto generated tracks are flagged via isAutogenerated: true.

How does the most replayed feature work?

YouTube ships an internal heatmap with intensityScoreNormalized values from 0 to 1 across video segments. The actor returns every segment so you can find the highest engagement moments programmatically. Useful for clip generation, viewer retention analysis, and content research.

Can I scrape comments?

Yes. Toggle on extractTopComments and the actor scrolls the comment thread and returns the top N comments. Each comment ships with author, channel handle, text, like count, posted time, pinned and hearted flags, and reply count. Toggle on extractCommentReplies for the first batch of replies per thread.

How do you handle the YouTube consent wall?

The actor sets CONSENT=YES+1 and SOCS=CAI cookies before navigation, which dismisses the EU consent banner without a click. If a consent dialog still renders, the actor clicks the accept button automatically.

What if a video is age gated?

Age gated videos require a signed-in session. The actor logs a warning and skips them. For pre-shipping content audits, age gating is a small minority of cases.

Can I filter by upload date and duration in search?

Yes. Pass uploadDate: "month" and duration: "long" and the actor builds the YouTube sp= filter parameter. Combine with sortBy: "viewCount" to find the highest performing month-old long form videos in any niche.

Does it work for shorts?

Yes. Shorts are detected automatically and tagged with type: "short" and flags.isShort: true. The row carries both url and shortsUrl. Pass maxShortsPerSearch to control how many shorts come back per query.

Is scraping YouTube allowed?

This actor reads HTML any anonymous viewer can see. Do not redistribute video files or transcripts you have no lawful basis to publish. Respect YouTube's terms.

---

Related actors

• TikTok Scraper. Videos, profiles, hashtags, and music data with optional video file and subtitle downloads.
• Instagram Scraper. Posts, reels, stories, comments, and follower data.
• LinkedIn Jobs Scraper Pro. Search URL, company URL, and recruiter contact extraction.
• Website Content Crawler. Websites to clean Markdown with token counts and RAG ready chunks.
• Google Maps Scraper. Local business data with reviews.
• Amazon Product Scraper. Product details, BSR, variants, ratings histogram across 10 marketplaces.