YouTube Scraper - Videos, Channels & Search (No API Key)

Extract public YouTube data at scale without the YouTube Data API or its daily quotas. Give it video URLs, channels, playlists or search queries and get clean, structured metadata for every video — plus optional LLM-ready transcripts.

What it does

• Videos — title, description, view count, like count, comment count, duration, publish/upload date, category, keywords, hashtags, live/short flags, and full thumbnail set.
• Channels — a channel-level record with name, subscriber count, video count, total views, description and external links.
• Discovery — expand channels, playlists and search queries into their videos (capped by maxVideosPerSource).
• Transcripts (optional add-on) — clean caption text per video, with language preference and optional translation.

Inputs

Field	Type	Notes
videoUrls	array	Video URLs or 11-char IDs (watch, youtu.be, Shorts).
channelUrls	array	@handle or /channel/UC... URLs. Pulls recent uploads.
playlistUrls	array	playlist?list=... URLs.
searchQueries	array	Scrapes the top matching videos per query.
videoType	select	videos (default), shorts, or streams — which channel tab to expand.
maxVideosPerSource	int	Cap per channel/playlist/search. Default 25.
includeChannelInfo	bool	Emit a channel record per channel URL. Default true.
includeTranscript	bool	Fetch transcript per video (add-on). Default false.
languages / translateTo	array / string	Transcript language preference & optional translation.
proxyConfiguration	proxy	Residential proxy strongly recommended — YouTube blocks datacenter IPs.

Output

One dataset item per video (and one per channel when includeChannelInfo is on). Example video item:

{
  "videoId": "dQw4w9WgXcQ",
  "url": "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
  "title": "…",
  "channel": "…",
  "channelId": "UC…",
  "viewCount": 1600000000,
  "likeCount": 18000000,
  "commentCount": 2100000,
  "durationSeconds": 213,
  "publishDate": "2009-10-25",
  "category": "Music",
  "keywords": ["…"],
  "hashtags": ["…"],
  "isShort": false,
  "isLive": false,
  "thumbnail": "https://i.ytimg.com/…"
}

Use cases

• Competitor and channel analysis (upload cadence, engagement, catalog size)
• Trend and topic discovery via search
• Influencer/creator vetting
• Feeding video metadata + transcripts into RAG / LLM pipelines

Field availability & notes

Verified live (Sept 2026):

• Videos — reliable: title, description, channel, channelId/URL, viewCount, likeCount, durationSeconds, publishDate, uploadDate, category, keywords, hashtags, isLive, isShort, isFamilySafe, thumbnails.
• Videos — not in v1: commentCount is null. YouTube loads comment counts lazily via a separate request, so it isn't available from the initial page. (Matches how other scrapers treat comments as a separate paid step.)
• Transcripts — optional add-on (off by default): enable includeTranscript to also fetch clean caption text per video (with a fresh residential IP per request to avoid rate limits). Uses youtube-transcript-api 1.x.
• Channels — reliable: channelName, channelId, description, subscriberCount, isVerified. videoCount is approximate for large channels (YouTube shows it abbreviated, e.g. "1K videos").
• Channels — may be null in v1: total channel viewCount, country, and external links live on the lazily-loaded About panel and aren't in the static HTML.
• Engagement/structure fields are best-effort; YouTube changes its page layout often.
• Use a residential proxy for reliable results at any volume — YouTube blocks datacenter IPs.

Monetization (set in Apify Console → Publication → Monetization)

The code already emits pay-per-event charges. Define these events to enable them:

Event name	When it fires	Suggested price (free tier)
video-scraped	Each video successfully scraped	~$0.004
channel-scraped	Each channel-info record	~$0.004
transcript-scraped	Each video transcript (add-on)	~$0.005

Benchmarks: streamers/youtube-scraper charges ~$0.004/video; transcripts elsewhere run $0.048/min.