YouTube Scraper — Search, Videos, Channels & Comments

Search, videos, channels & comments in one Actor. No Google API key, no daily quota — just $0.25 per 1,000 results.

Scrape YouTube search results, video details, channel stats and comments as clean, structured data. Paste YouTube links or keywords — no Google API key, no daily quota. One Actor covers seven common tasks: pick a task, fill one box, hit Start.

• ✅ $0.25 per 1,000 results — pay only for rows you actually get
• ✅ No Google API key, no daily quota — nothing to set up
• ✅ Paste links, not IDs — full watch?v=, youtu.be/, /shorts/, @handle and /channel/UC… URLs all work
• ✅ Pagination handled for you — just say how many results you want
• ✅ Flat rows that export straight to CSV, Excel, Google Sheets or JSON
• ✅ Works from the UI and from the API — same input either way

---

What can it do?

Task	What you give it	What you get back
🔍 Search videos	Keywords	Matching videos with title, channel, publish date, thumbnail
🔍 Search channels	Keywords	Matching channels with title, description, thumbnail
🎬 Video details	Video URLs or IDs	Views, likes, comment count, duration, tags, category, captions
📺 Channel details	Channel URLs, @handles or IDs	Subscribers, total views, video count, country, handle
📺 Channel videos	Channel URLs, @handles or IDs	The channel's uploads, newest first
💬 Video comments	Video URLs or IDs	Top-level comments + author, likes, date (replies optional)
💬 Comment replies	Comment IDs	All replies to those comments

---

Quick start (UI)

1. Choose a task in “What do you want to get?”.
2. Fill in only the box that belongs to it — each box says which task it is for:
   • Search keywords → for the two search tasks
   • Videos (URLs or IDs) → for video details and video comments
   • Channels (URLs, handles or IDs) → for channel details and channel videos
   • Comment IDs → for comment replies
3. Optionally set Max results per input, a country, a date range or a sort order.
4. Click Start. Results appear on the Storage → Dataset tab, ready to export.

Every box comes pre-filled with a working example, so you can switch the task dropdown and hit Start straight away to see what each one returns.

---

Examples

Search for videos

{
  "operation": "search_videos",
  "queries": ["coke studio", "lofi hip hop"],
  "order": "viewCount",
  "regionCode": "IN",
  "maxItems": 100
}

Get details for a few videos

{
  "operation": "video_details",
  "videos": [
    "https://www.youtube.com/watch?v=4zTFzMPWGLs",
    "https://youtu.be/dQw4w9WgXcQ",
    "https://www.youtube.com/shorts/-2Wqk2ncV7c"
  ]
}

Get channel stats from a handle

{
  "operation": "channel_details",
  "channels": ["https://www.youtube.com/@MrBeast", "@mkbhd"]
}

List a channel's newest uploads

{
  "operation": "channel_videos",
  "channels": ["UCu59yAFE8fM0sVNTipR4edw"],
  "order": "date",
  "publishedAfter": "2025-01-01",
  "maxItems": 200
}

Scrape comments (with replies)

{
  "operation": "comment_threads",
  "videos": ["https://www.youtube.com/watch?v=32mJgqhnbGY"],
  "commentOrder": "relevance",
  "includeReplies": true,
  "maxRepliesPerComment": 50,
  "maxItems": 500
}

---

Output

Rows are flat and consistent, so CSV and Sheets exports just work.

Videos (search / channel videos)

{
  "videoId": "5Eqb_-j3FDA",
  "title": "Pasoori | Coke Studio Season 14",
  "url": "https://www.youtube.com/watch?v=5Eqb_-j3FDA",
  "publishedAt": "2022-02-07T13:30:11Z",
  "channelTitle": "Coke Studio Pakistan",
  "channelId": "UCM1VesJtJ9vTXcMLLr_FfdQ",
  "channelUrl": "https://www.youtube.com/channel/UCM1VesJtJ9vTXcMLLr_FfdQ",
  "description": "Let's transcend boundaries…",
  "thumbnailUrl": "https://i.ytimg.com/vi/5Eqb_-j3FDA/hqdefault.jpg",
  "liveBroadcastContent": "none",
  "searchQuery": "coke studio"
}

Video details adds viewCount, likeCount, commentCount, duration, durationSeconds, tags, categoryId, topicCategories, definition, hasCaptions, privacyStatus, madeForKids, defaultAudioLanguage.

Channel details

{
  "channelId": "UCFFbwnve3yF62-tVXkTyHqg",
  "title": "Zee Music Company",
  "handle": "@zeemusiccompany",
  "url": "https://www.youtube.com/@zeemusiccompany",
  "subscriberCount": 122000000,
  "viewCount": 89211450788,
  "videoCount": 16576,
  "publishedAt": "2014-03-12T11:59:25Z",
  "country": "IN",
  "uploadsPlaylistId": "UUFFbwnve3yF62-tVXkTyHqg"
}

Comments

{
  "commentId": "UgwOSGaPuVONHxyrtxV4AaABAg",
  "text": "does he speak Kannada ?",
  "authorDisplayName": "@prashanthalex7702",
  "likeCount": 12,
  "publishedAt": "2025-03-18T23:26:33Z",
  "isReply": false,
  "parentId": null,
  "totalReplyCount": 1,
  "videoId": "32mJgqhnbGY",
  "videoUrl": "https://www.youtube.com/watch?v=32mJgqhnbGY",
  "authorChannelId": "UCWS904BG8MIrnbCMj8qiNfQ"
}

Replies come as extra rows with isReply: true and parentId set to the comment they belong to.

Counts are real numbers (not strings), so you can sort and total them right away. Need the untouched nested objects instead? Turn on Raw output (advanced).

---

Using it from the API

Run the Actor and get results in one call:

curl -X POST "https://api.apify.com/v2/acts/trysmartapi~youtube-scraper/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{
    "operation": "search_videos",
    "queries": ["coke studio"],
    "maxItems": 50
  }'

Same thing in JavaScript:

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: 'YOUR_APIFY_TOKEN' });

const run = await client.actor('trysmartapi/youtube-scraper').call({
    operation: 'comment_threads',
    videos: ['https://www.youtube.com/watch?v=32mJgqhnbGY'],
    maxItems: 200,
    includeReplies: true,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

And in Python:

from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")

run = client.actor("trysmartapi/youtube-scraper").call(run_input={
    "operation": "channel_details",
    "channels": ["@MrBeast", "@mkbhd"],
})

for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item)

---

Input reference

Field	Type	Used by	Notes
operation	string	all	search_videos, search_channels, video_details, channel_details, channel_videos, comment_threads, comment_replies
queries	string[]	search tasks	One keyword per entry; each is searched separately
videos	string[]	video_details, comment_threads	Video URLs or 11-char IDs
channels	string[]	channel_details, channel_videos	Channel URLs, @handles or UC… IDs
commentIds	string[]	comment_replies	Comment IDs from a comments run
maxItems	integer	listing tasks	Cap per keyword / video / channel. Default 100
includeReplies	boolean	comment_threads	Adds reply rows. Default false
maxRepliesPerComment	integer	comment_threads	Default 50
commentOrder	string	comment_threads	time (newest) or relevance
order	string	search / channel videos	relevance, date, viewCount, rating, title. Use date on channels for full history
videoDuration	string	video listings	any, short (<4 m), medium (4–20 m), long (>20 m)
regionCode	string	listings	ISO country code, e.g. US, IN, GB
videoCategoryId	string	video listings	e.g. 10 Music, 20 Gaming, 24 Entertainment
publishedAfter / publishedBefore	string	listings	YYYY-MM-DD or full ISO date-time
eventType	string	search_videos	any, live, upcoming, completed
rawOutput	boolean	all	Return nested API objects instead of flat rows
startToken	string	listings	Resume a previous run (single input item only)

---

Notes & tips

• maxItems is per input. Three keywords with maxItems: 100 returns up to 300 rows.
• Resuming a run: each run stores a nextToken per input in its key-value store under OUTPUT. Paste one into startToken to continue from the next page.
• Video and channel details are batched — asking for 200 videos costs a handful of requests, not 200.
• Channel videos walks the full upload history. YouTube's channel listing only exposes the newest ~15 uploads, so the Actor slices the request into one-year windows and pages through each, newest year first. Set maxItems as high as you need. Sorting by anything other than "Newest first" uses the plain listing instead and returns far fewer videos — use "Newest first" when you want depth.
• Private, deleted or region-blocked items are skipped and reported as run-log warnings.
• @handles are resolved automatically to channel IDs, which costs one extra lookup. Pass UC… IDs directly if you already have them.
• Comment counts on a video can be larger than what is fetchable — YouTube hides some comments and disables them entirely on some videos.

---

Pricing

$0.25 per 1,000 results — charged per row written to the dataset, nothing else. No monthly fee, no minimum, no charge for rows you don't receive.

You scrape	You pay
100 results	$0.03
1,000 results	$0.25
10,000 results	$2.50
100,000 results	$25.00

Both a comment and a video row count as one result. Use Max results per input to keep any run inside a budget you choose.

---

Legal

This Actor retrieves publicly available YouTube data only. It does not access private data or bypass any login. You are responsible for using the collected data in line with YouTube's terms and applicable law (including GDPR when handling personal data such as comment author names).