YouTube Scraper & API: Videos, Channels, Search, Comments avatar

YouTube Scraper & API: Videos, Channels, Search, Comments

Pricing

from $1.25 / 1,000 results

Go to Apify Store
YouTube Scraper & API: Videos, Channels, Search, Comments

YouTube Scraper & API: Videos, Channels, Search, Comments

YouTube videos, channels, Shorts, playlists and search results plus full comment threads: views, likes, subscribers, publish dates, chapters and description links. Or fetch one video via API.

Pricing

from $1.25 / 1,000 results

Rating

0.0

(0)

Developer

Philip Kirkbride

Philip Kirkbride

Maintained by Community

Actor stats

0

Bookmarked

10

Total users

7

Monthly active users

12 hours ago

Last modified

Features

Limited permissionsSingle-tenantServer interface

Share

YouTube Scraper: Videos, Channels, Search & Comments

Search results, channels, comments with replies, Shorts and playlists from YouTube, all in one run. Paste any YouTube link, @handle or search term to collect video, channel and comment rows with views, likes, subscribers, exact durations, publish dates, descriptions, hashtags and full comment threads. No login, API key or quota. Export to CSV, Excel or JSON, or send the rows straight to Google Sheets.

Built for creator and competitor research, content and SEO research, brand and sentiment monitoring, and campaign reporting.

  • Collect everything in one run, including videos, channels with their About details, search results with YouTube's own filters, Shorts, live streams, public playlists, comments and replies. A video can appear once per matching term, with its own searchRank when ranking options are used.
  • Rich video rows: chapters, every link in the description (sponsors, merch, socials, other videos), the verified badge, hashtags, likes, comment counts and exact publish dates.
  • Clean, honest data: missing values are empty, never a made-up zero, and the run summary names any link that could not be collected.
  • Low price per row, the same for videos, channels and comments, with a hard row cap so you control the spend.

Switching from another YouTube scraper? Paste your usual startUrls / searchQueries input: this Actor accepts the common startUrls / searchQueries input format, including keywords, youtubeHandles, sortVideosBy, oldestPostDate, maxResultStreams and maxResultsShorts.

Need transcripts? Use the companion YouTube Transcripts Scraper for timestamped captions of the same videos.

Also available

The YouTube suite: pick the job.

ActorUse when
YouTube Scraper (this Actor)Videos, channels, search results and comments
YouTube TranscriptsBulk transcripts as text, SRT or VTT
YouTube Outlier FinderBreakout videos that beat their channel median
TikTok Profiles & VideosCreator stats and latest video metrics
Instagram ProfilesFollowers, bio and latest posts
Twitch Channels & StreamsLive streams, channels, VODs and clips

Quick Start

{
"videoUrls": ["https://www.youtube.com/watch?v=dQw4w9WgXcQ"],
"channelUrls": ["https://www.youtube.com/@mkbhd"],
"searchQueries": ["python tutorial"],
"maxResults": 10,
"maxComments": 50
}

This run returns the video's full statistics plus its top 50 comments, the MKBHD channel profile plus its 10 latest uploads, and the first 10 videos YouTube shows for "python tutorial". The Console's sample input is a smaller version of the same mix, so the first run needs no edits.

Paste any YouTube link. Put video, Shorts, live, channel, channel tab (/videos, /shorts, /streams), playlist, search-result and hashtag links into startUrls, and each one is collected as what it is:

{
"startUrls": [
{"url": "https://www.youtube.com/@mkbhd/shorts"},
{"url": "https://www.youtube.com/results?search_query=python+tutorial&sp=EgIIAw%3D%3D"},
{"url": "https://www.youtube.com/playlist?list=PLNcg_FV9n7qZGfFl2ANI_zISzNp257Lwn"}
],
"maxResults": 20
}

Use cases

  • Creator and competitor research: subscribers, total views, upload cadence and per-video performance for any set of channels, sorted by latest, most popular or oldest, or limited to recent uploads.
  • Content and SEO research: which videos and channels rank for a search term in a given country, filtered by upload date, length or features such as 4K or live.
  • Audience and sentiment analysis: comments and replies with likes, reply counts and creator hearts, for chosen videos or every video a search or channel returns.
  • Campaign reporting: refresh views, likes and comment counts for a list of sponsored or owned videos.
  • Course and catalog building: public playlists in order, with each video's position.

Input

Sources (use any mix):

InputMeaning
startUrlsAny YouTube links: videos, Shorts, live streams, channels and their /videos, /shorts or /streams tabs, public playlists, search-result pages (their filters are kept) and hashtag pages.
videoUrlsWatch, youtu.be, Shorts or live links, or bare 11-character video IDs. Each becomes one full video row; Shorts links are marked isShort: true.
channelUrlsChannel links, @handles or UC… channel IDs. Each adds one channel profile row plus its latest uploads.
searchQueriesSearch terms. Results come in YouTube's order. searchTerms is an older name for the same list.
playlistUrlsPublic playlist links or bare PL… list IDs. Watch links with list=PL… collect from the beginning of the playlist.

A link pasted into the wrong list is collected as what it is (for example a channel link in videoUrls), and OUTPUT.notices says so.

Search filters (all optional; empty means YouTube's normal results): searchSort (relevance or popularity), uploadDate (today, week, month, year), duration (short under 3 minutes, medium 3–20 minutes, long over 20 minutes), searchType (video, shorts or channel; channel searches return channel rows), searchFeatures (live, 4k, 360, vr180, creative_commons), includeShorts (default true), country (the search market, two-letter code, default US) and language (the language of the search results, default English).

Only search uses the requested language. Every other page is read in English, including video details, channel and playlist pages, and comments. Titles, descriptions and labels from those pages reflect what YouTube shows English-language viewers, including translations where available. On localized search cards, subscriber counts may be missing and localized live badges may leave isLive false. Listing counts can be rounded in any language.

Channel options: includeChannelInfo (profile row, default true), channelSort (latest, popular or oldest), publishedAfter (a date such as 2026-10-01 or a relative value such as 30 days), includeChannelShorts (also the Shorts tab) and includeChannelStreams (also the Live tab). For Shorts or live streams only, set maxResults to 0 and use a positive limit for that tab. Set includeChannelInfo: false for video rows only. A /shorts or /streams link in startUrls uses maxResults by default. Its tab-specific limit overrides that budget.

Set the Live limit with maxStreamsPerChannel or its JSON alias maxResultStreams. Set the Shorts limit in JSON with maxResultsShorts.

Comments: maxComments (per video, default 0), commentsSort (top or newest), includeReplies, and commentsForAllVideos to collect comments for every video found by searches, channels, playlists and hashtags as well.

Limits: maxResults (videos per search term, channel and playlist, default 20), maxItems (hard stop for the whole run, default 1000) and videoDetails (likes, comment count, exact publish date, subscribers and full description for search, channel and playlist videos).

Connection: proxyConfiguration. The default works for most runs; you can also use a direct connection, Apify datacenter proxy or your own proxy URLs.

What it returns

Every row has a type (video, channel or comment), the input that produced it and a scrapedAt timestamp.

Video rows

FieldMeaning
videoId, url, title, descriptionVideo identity and text. Search, channel and playlist rows carry the result-list snippet unless videoDetails is on.
viewCount, likeCount, commentCountLikes and comment count come with full details (video links or videoDetails). Search, channel, Shorts and playlist cards can show rounded view counts, such as 1.3M. Turn on videoDetails for the exact view count when available. Comment counts above 1,000 are rounded the way YouTube shows them.
publishedAt, publishedTextPublish date (YYYY-MM-DD) and the date as YouTube shows it, for example 2 days ago.
duration, durationTextLength in seconds and as m:ss / h:mm:ss.
channelId, channelName, channelUrl, channelHandle, subscriberCountThe uploader. Subscriber counts are YouTube's rounded figures.
hashtags, thumbnailUrl, isLive, isShortHashtags from the title area and description, a stable thumbnail link, live and Shorts flags.
detailfull for rows with full statistics, listing for search, channel and playlist result rows.
chaptersThe chapters the creator lists in the description, as {"startSeconds", "title"} in order; empty when the video has none. Full-detail rows.
descriptionLinksEvery link in the description as {"text", "url"}, in order and each once: sponsor, shop and social links (with their real destination), other videos and channels. Full-detail rows.
channelIsVerifiedtrue when the uploader has YouTube's verified or official-artist badge. Full-detail rows.
searchQuery, searchRankSearch rows carry the search term. searchRank is the one-based position of the first card for the result, in YouTube's order. Excluded Shorts and repeated cards occupy their original positions. Rank appears when you use a search filter, country, language or a filtered search link. A matching video can appear once per term. Channel searches (searchType: channel) return each matching channel at most once, under the first term that matched it. Video links and channels are processed first and take precedence over search terms, which take precedence over playlists and hashtags.
playlistId, playlistPositionPresent on playlist-sourced videos, including full-detail rows. Position is one-based and null when YouTube omits it.

Channel rows: channelId, channelName, channelUrl, channelHandle, description, subscriberCount, subscriberCountText, videoCount, viewCount (total channel views), joinedDate, country, links, keywords, avatarUrl, isFamilySafe and aboutComplete. Channel rows from a searchType: channel search carry what the result card shows plus searchQuery and searchRank; the rest is null or empty and aboutComplete is false.

Comment rows: commentId, videoId, videoUrl, text, authorName, authorChannelId, authorIsChannelOwner, authorIsVerified, likeCount (rounded the way YouTube shows it), replyCount, publishedText, isEdited, isPinned, isReply and parentCommentId. With commentsForAllVideos, comment rows also carry hasCreatorHeart.

The OUTPUT record of the run's key-value store summarizes the run: row counts by type, the run status, notices explaining anything the run adjusted, and an errors list naming each input that could not be collected, with a code and a readable message. Playlist runs also include OUTPUT.playlists with delivered, duplicate and unavailable counts and the stop reason.

Output example

Illustrative rows (live values change constantly):

{"type": "video", "input": "https://www.youtube.com/watch?v=dQw4w9WgXcQ", "scrapedAt": "2026-10-19T22:36:59Z", "videoId": "dQw4w9WgXcQ", "url": "https://www.youtube.com/watch?v=dQw4w9WgXcQ", "title": "Rick Astley - Never Gonna Give You Up (Official Video) (4K Remaster)", "description": "The official video for “Never Gonna Give You Up” by Rick Astley. ...", "viewCount": 1828414467, "likeCount": 19489750, "commentCount": 2400000, "publishedAt": "2009-10-24", "publishedText": "Oct 24, 2009", "duration": 214, "durationText": "3:34", "channelId": "UCuAXFkgsw1L7xaCfnd5JJOw", "channelName": "Rick Astley", "channelUrl": "https://www.youtube.com/@RickAstleyYT", "channelHandle": "@RickAstleyYT", "subscriberCount": 4550000, "hashtags": ["RickAstley", "NeverGonnaGiveYouUp"], "thumbnailUrl": "https://i.ytimg.com/vi/dQw4w9WgXcQ/hqdefault.jpg", "isLive": false, "isShort": null, "detail": "full", "descriptionLinks": [{"text": "https://linktr.ee/rickastleynever", "url": "https://linktr.ee/rickastleynever"}], "chapters": [], "channelIsVerified": true}
{"type": "channel", "input": "https://www.youtube.com/@mkbhd", "scrapedAt": "2026-10-19T22:37:02Z", "channelId": "UCBJycsmduvYEL83R_U4JriQ", "channelName": "Marques Brownlee", "channelUrl": "https://www.youtube.com/@mkbhd", "channelHandle": "@mkbhd", "description": "MKBHD: Quality Tech Videos | YouTuber | Geek | Consumer Electronics | Tech Head | Internet Personality!", "subscriberCount": 21400000, "subscriberCountText": "21.4M subscribers", "videoCount": 1859, "viewCount": 5798311104, "joinedDate": "2008-03-21", "country": "United States", "links": [{"title": "Twitter", "url": "https://twitter.com/MKBHD"}], "keywords": ["MKBHD", "MarquesBrownlee"], "avatarUrl": "https://yt3.googleusercontent.com/qu4TmIaYUlS41-dJ9gZ7DUR3nilvmB5_11i6OKSdvNnBNiyOusZP1bMN6ICnuxtjFBb6ioKgRQ=s900-c-k-c0x00ffffff-no-rj", "isFamilySafe": true, "aboutComplete": true}
{"type": "video", "input": "python tutorial", "scrapedAt": "2026-10-19T22:37:05Z", "videoId": "fWjsdhR3z3c", "url": "https://www.youtube.com/watch?v=fWjsdhR3z3c", "title": "Learn Python in Less than 10 Minutes for Beginners (Fast & Easy)", "description": "In this crash course I'll be teaching you the basics of Python in less than 10 minutes. ...", "viewCount": 1843779, "likeCount": null, "commentCount": null, "publishedAt": null, "publishedText": "5 years ago", "duration": 630, "durationText": "10:30", "channelId": "UCuudpdbKmQWq2PPzYgVCWlA", "channelName": "Indently", "channelUrl": "https://www.youtube.com/@Indently", "channelHandle": "@Indently", "subscriberCount": null, "hashtags": [], "thumbnailUrl": "https://i.ytimg.com/vi/fWjsdhR3z3c/hqdefault.jpg", "isLive": false, "isShort": false, "detail": "listing", "searchQuery": "python tutorial"}
{"type": "comment", "input": "https://www.youtube.com/watch?v=dQw4w9WgXcQ", "scrapedAt": "2026-10-19T22:37:08Z", "commentId": "UgzKYksqeQ5gA3t0g0F4AaABAg", "videoId": "dQw4w9WgXcQ", "videoUrl": "https://www.youtube.com/watch?v=dQw4w9WgXcQ", "text": "I Rick rolled my entire French class ...", "authorName": "@grustache4569", "authorChannelId": "UCKrzgYonJnMd0Dk7zQXjYXw", "authorIsChannelOwner": false, "authorIsVerified": false, "likeCount": 104000, "replyCount": 623, "publishedText": "6 years ago", "isEdited": true, "isPinned": false, "isReply": false, "parentCommentId": null}

FAQ

Can I mix videos, channels, searches and playlists in one run? Yes. Fill any combination of the source lists, or put every link into startUrls. A video delivered from two different sources arrives once; a video that ranks for two of your search terms arrives once per term, and each of those rows is billed like any other row (searchRank appears when a search filter, country, language or a filtered search link was used). A video already delivered from a video link or a channel does not repeat in search, playlist or hashtag rows.

How do I get only recent channel videos? Set publishedAfter, for example 30 days. Without videoDetails it goes by the age YouTube shows, so a video posted just before the cutoff can slip in; turn on videoDetails for exact dates.

How do I get comments for videos found by a search or on a channel? Set maxComments and turn on commentsForAllVideos. Each video's comments follow its row, in one run.

How do I get exact publish dates, likes and comment counts? Turn on videoDetails. Individual video links always return full details.

How do I get chapters and the links in a video's description? They come with full details: always for video links, and for search, channel and playlist videos when videoDetails is on. Use descriptionLinks to find sponsors, affiliate and merch links or a creator's social accounts.

How do I collect a channel's Shorts or live streams? Turn on includeChannelShorts or includeChannelStreams to collect the tab alongside uploads. Set maxResults to 0 and use a positive maxResultsShorts or maxResultStreams to collect only the tab. A /shorts or /streams link in startUrls uses maxResults by default. The corresponding maxResultsShorts or maxResultStreams overrides that budget. Set includeChannelInfo: false for video rows only.

Why are some counts missing or rounded? YouTube rounds subscriber counts, comment counts above 1,000 and comment likes. Search, channel, Shorts and playlist cards can also show rounded view counts. Turn on videoDetails for the exact view count when available. Missing values are null and a real zero stays zero.

Why did my run collect one video I did not ask for? An input with no sources runs the Store example video so the form still returns a result. OUTPUT.usedExample is true and the status line says so, naming any input fields that were not recognized.

How do I keep the cost predictable? Set maxItems. The run stops as soon as that many rows have been delivered.

Can I run it on a schedule or from code? Yes. Schedule runs on Apify or start them with the Apify API, then read each run's dataset.

Use it as an API

Get a YouTube video, channel, or set of search results on demand. Get the answer in JSON from one HTTPS call. Call it from an AI agent through an HTTP-capable MCP tool, or add an HTTP request step in n8n, Make, or Zapier.

curl --get 'https://automatic--youtube-scraper.apify.actor/' \
--header "Authorization: Bearer $APIFY_TOKEN" \
--data-urlencode 'url=https://www.youtube.com/watch?v=dQw4w9WgXcQ'

Send the same fields as a JSON object with POST /. Results are in data.items and in the run's default dataset. See GET /openapi.json for the request and response definitions.

Price

$0.00125 per delivered row (video, channel or comment), plus $0.005 per run start. A start is charged once per GB of memory, so the default 512 MB run is one start. Inputs that fail are not dataset rows and incur no row charge; the run-start charge still applies. Platform usage is included in these prices.

Good to know

  • Public availability varies. Hidden, private, removed, members-only, age-restricted or region-restricted entries may be missing; OUTPUT.errors names each input that could not be collected. A run where every input was unusable or empty (a mistyped handle, a removed video, a search with no results) ends with no rows and a status message saying why.
  • Links pasted into the wrong list are collected as what they are, and OUTPUT.notices names each one. Unrecognized input fields are ignored and listed there too.
  • Residential connections are not used. A request for one runs on the standard connection, with a notice.
  • Without videoDetails, publishedAfter uses the upload age YouTube shows (such as "3 weeks ago"), read as a minimum: "1 month ago" counts as at least four weeks old. No recent video is dropped, but a few uploaded just before the cutoff can be kept, so it is approximate; publishedAt stays empty rather than estimated. Shorts listings show no upload age and are kept unless videoDetails is on.
  • searchFeatures covers the filters that YouTube reliably applies: live, 4K, 360°, VR180 and Creative Commons. A hashtag page shows its top videos only (about 20). Shorts shown inside a search's Shorts shelf carry only the title and view count on the listing row; turn on videoDetails for their channel, date and length. Search results keep YouTube's on-screen order, so a page's Shorts count towards maxResults.
  • Playlist collection stops at the requested caps or after at most 100 pages or 240 seconds per playlist; it does not promise a complete playlist. Only public PL… lists are supported.
  • Listing fields reflect what YouTube publishes; descriptions, likes and exact dates may be null until videoDetails is enabled. Uploader keyword tags are unavailable; hashtags contains public hashtags. Captions and downloads are not included.
  • chapters lists the chapters a creator writes into the description (starting at 0:00, three or more). Chapters YouTube generates automatically are not included.
  • Use public data only where you have the necessary permissions and a lawful basis; YouTube's terms restrict automated access.

Your feedback

Found a bug, or need a field or filter we don't have yet? Open an issue on this Actor's Issues tab. We read every issue and use them to decide what to build next.