YouTube Scraper — Videos, Transcripts, Comments, Channels & Search

Scrape YouTube at scale into clean JSON, CSV, Excel or XML. This YouTube scraper extracts video metadata, full transcripts, comments with reply threads, channel profiles, playlists and search results — all from one Actor, with no API key and no quotas. A faster, lower-cost alternative to the official YouTube Data API.

Most YouTube scrapers make you run three or four separate tools — one for metadata, one for transcripts, one for comments — and bill you for each. This one does it all in a single run.

What does the YouTube Scraper do?

The YouTube Scraper crawls and extracts public YouTube data by video URL, playlist URL, channel URL or search term. Point it at a video, a Shorts link, a playlist, a @handle or channel, or a search query, and it returns structured data points — views, likes, duration, publish date, subscriber counts, transcripts, comments and more — ready for analytics, datasets, LLM pipelines or lead generation.

What data can you scrape from YouTube?

Group	Fields
🎥 Video	id, title, description, description links, view count, like count, duration (+ seconds), publish date, keywords/tags, thumbnails, is-monetized, comments-turned-off, is-Short, is-Live, family-safe flag, category
📝 Transcript	full text, timestamped segments (start + duration), ready-to-use SRT subtitles, language, auto-generated flag, list of available languages
💬 Comments	comment text, author handle, like count, published time, reply threads, creator/verified flags, reply count
📺 Channel	name, handle, channel ID, subscriber count, total views, video count, country, join date, avatar, description, external & social links
🔍 Search	every matching video with full metadata, ranked by YouTube relevance

Use cases

• Feed transcripts to an LLM — summaries, RAG knowledge bases, content repurposing and translation at scale.
• Comment & sentiment analysis — mine top comments and reply threads for any video or channel.
• Influencer & lead generation — pull subscriber counts, contact emails and social links from channel profiles.
• Competitor & trend research — track a niche via search results, top videos and channel catalogs.
• Brand & content monitoring — watch mentions, engagement and new uploads over time.
• Build a video dataset — views, likes, duration and metadata for thousands of videos.
• Content strategy — find high-performing formats, keywords and titles in your space.
• Academic & market research — study engagement, topics and audience signals from public data.

How to scrape YouTube data

1. Click Try for free / Start to open the YouTube Scraper.
2. Paste video URLs, playlist URLs, a search query, or channel URLs (bare video IDs and @handles also work).
3. Toggle the add-ons you want — transcripts, comments, and channel enrichment — and set limits.
4. Click Run.
5. Download the results as JSON, CSV, Excel or XML, or pull them from the API.

Input

Provide any mix of video URLs, search queries and channel URLs, then toggle the add-ons.

{
  "videoUrls": ["https://www.youtube.com/watch?v=aircAruvnKk"],
  "playlistUrls": ["https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi"],
  "searchQueries": ["machine learning tutorial"],
  "maxResultsPerQuery": 20,
  "channelUrls": ["https://www.youtube.com/@3blue1brown"],
  "includeTranscript": true,
  "transcriptLanguage": "en",
  "includeComments": true,
  "maxComments": 100,
  "includeChannel": true
}

Field	Type	Description
videoUrls	array	YouTube video or Shorts URLs (or bare 11-char video IDs).
playlistUrls	array	Playlist URLs (or IDs) — every video in the playlist is scraped with full metadata.
searchQueries	array	Search terms; each result is scraped as a video.
maxResultsPerQuery	integer	Videos returned per search query, and per playlist (default 20).
channelUrls	array	Channel URLs — @handle or /channel/UC….
includeTranscript	boolean	Attach the full transcript (text, segments, SRT) to each video.
transcriptLanguage	string	Preferred transcript language code, e.g. en, es, de.
includeComments	boolean	Attach comments with reply threads to each video.
maxComments	integer	Max comments per video, including replies.
includeChannel	boolean	Attach the uploader's channel profile to each video.

Scrape YouTube by URL or search term

Pass exact videoUrls when you know the videos, or searchQueries to discover them — mix both in one run.

Scrape YouTube transcripts and subtitles

Enable includeTranscript for the full transcript, timestamped segments and downloadable SRT subtitles, with transcriptLanguage selection and an auto-generated fallback.

Scrape YouTube comments

Enable includeComments to pull top comments including reply threads, each with author, like count and verified/creator flags.

Output

Each video is one dataset item:

{
  "type": "video",
  "id": "aircAruvnKk",
  "title": "But what is a neural network? | Deep learning chapter 1",
  "viewCount": 23621207,
  "likeCount": 551760,
  "commentCount": 8300,
  "durationSeconds": 1120,
  "publishDate": "2017-10-05T08:11:25-07:00",
  "descriptionLinks": ["https://www.patreon.com/3blue1brown", "https://3b1b.co/nn1-thanks"],
  "isMonetized": true,
  "commentsTurnedOff": false,
  "keywords": ["neural network", "deep learning"],
  "channel": { "name": "3Blue1Brown", "handle": "@3blue1brown", "subscriberCount": 8460000, "links": ["…"] },
  "transcript": {
    "available": true,
    "language": "en",
    "segments": [ { "start": 4.2, "duration": 3.1, "text": "This is a 3." } ],
    "fullText": "This is a 3. It's sloppily written …",
    "srt": "1\n00:00:04,200 --> 00:00:07,300\nThis is a 3.\n…"
  },
  "comments": [
    { "author": "@user", "text": "Great video!", "likeCount": 13000, "replies": ["…"] }
  ]
}

Channels and standalone search results are emitted as type: "channel" and type: "search-result" items with their own fields.

Output schema (video items)

Field	Type	Description
type	string	Item type — video, channel or search-result.
id	string	YouTube video ID.
url	string (URL)	Canonical watch URL of the video.
title	string	Video title.
description	string	Full video description text.
descriptionLinks	array	Links found in the description.
durationSeconds	integer	Video length in seconds.
viewCount	integer	Total view count.
likeCount	integer	Total like count.
commentCount	string	Number of comments (as shown by YouTube).
commentsTurnedOff	boolean	Whether comments are disabled on the video.
isMonetized	boolean	Whether the video appears to be monetized.
publishDate	string (ISO date)	When the video was published.
uploadDate	string (ISO date)	When the video was uploaded.
category	string	YouTube category, e.g. Education.
keywords	array	Video tags/keywords.
isLive	boolean	Whether the item is a live stream.
thumbnail	string (URL)	Thumbnail image URL.
channel	object	Channel info — name, handle, subscriberCount and links.
availableTranscripts	array	Transcript languages available for the video.

When enabled, video items also include a transcript object (segments, fullText, srt) and a comments array (author, text, likeCount, replies).

Need YouTube comments, transcripts or other platforms?

This Actor already bundles comments, transcripts and channel data — but if you're building a wider dataset, pair it with our other scrapers:

• TVmaze Scraper — TV shows, episodes and cast with IMDb IDs and ratings.
• Spotify Scraper — tracks, artists, albums and playlists.
• SoundCloud Scraper — tracks, artists and play counts.
• Twitch Scraper — streams, channels and viewer stats.
• Apple Podcasts Scraper — podcast shows and episodes.

FAQ

Do I need a YouTube API key? No. The YouTube Scraper needs no official API key and has no quotas — just add your inputs and run.

How many videos can I scrape per run? As many as you like. Set maxResultsPerQuery per search, and pass any mix of video URLs, search queries and channel URLs in one run to scale up to thousands.

Can I scrape YouTube transcripts and subtitles? Yes — enable includeTranscript for the full text, timestamped segments and ready-to-use SRT subtitles, with language selection.

Can I scrape YouTube comments? Yes — enable includeComments for top comments including reply threads, with author, like count and verified flags.

Can I scrape a whole YouTube playlist? Yes — pass playlist URLs in playlistUrls and every video in the playlist is scraped with full metadata (and transcripts/comments if enabled). Use maxResultsPerQuery to cap how many videos per playlist.

Can I download videos or audio? No. This Actor extracts metadata, transcripts, comments and channel data — text and numbers only, not media files.

What export formats are supported? JSON, CSV, Excel and XML, plus webhooks and the Apify API.

Can I automate or integrate the YouTube Scraper? Yes. The YouTube Scraper can be connected with almost any cloud service or web app thanks to integrations on the Apify platform. It works with Make, Zapier, Slack, Airbyte, GitHub, Google Drive and many more, plus the Apify API, JavaScript/Python clients and MCP. Or use webhooks to trigger an action whenever a run finishes — get a notification, or kick off another process such as loading your data downstream.

Is scraping YouTube legal? The Actor collects only public data. You are responsible for how you use it and for complying with YouTube's terms and applicable laws.

Related Actors

• Reddit Scraper — posts, comments and subreddits.
• Google Trends Scraper — interest over time for any topic.
• Google News Scraper — headlines and articles by query.
• Steam Games Scraper — game prices, genres and reviews.

Notes

Original clean-room implementation. Collects only public YouTube data; you are responsible for how you use the data and for complying with YouTube's terms. This is an independent tool and is not affiliated with or endorsed by YouTube.