YouTube Scraper: Videos, Shorts, Channels & Comments
The YouTube Scraper returns YouTube data as JSON rows: videos and Shorts from keyword search, channel tabs or playlists, channel rows with subscribers and About links, full video details with exact likes, publish date and description, and comments with replies. It calls YouTube over plain HTTPS, so there is no API key, login or browser. Every delivered row costs $1.00 per 1,000, with platform usage included. Run it from Apify Console, the API or MCP.
from $0.001 per result, $1.00 per 1,000 · Updated Nov 2026, version 0.1.15
Key facts
- Price
- from $0.001 per result, $1.00 per 1,000
- Pricing model
- Pay per event
- Output
- JSON, CSV, Excel, XML, RSS via Apify datasets
- Access
- Apify Console, REST API, Python and JS clients, MCP tool youtube
- Version
- Updated Nov 2026, version 0.1.15
- Usage
- 93.4% successful runs (30 days)
- Actor
subimpact/youtube-scraper
Who it is for
- Content teams tracking a niche's top videos by views and upload date
- Brands monitoring hundreds of channels for new uploads with an RSS feed
- Agencies auditing comment sentiment on campaign videos
- Researchers measuring channel growth from subscriber and lifetime view counts
Use cases
Rank a keyword's top videos
Put phrases in searchQueries, set searchSortBy to views and searchUploadDate to year, then cap with maxResultsPerQuery. Rows return position, videoId, viewCount, durationSeconds, publishedText, channelName and thumbnailUrl. Set searchIncludeShorts to true if you want Shorts shelves counted too, and read isShort to separate them from long form.
Monitor channels for new uploads
Set channelSource to rss and list handles or channel ids in startUrls. Each channel costs one small request and returns its 15 newest uploads. With onlyNewVideos on, delivered ids are stored in stateStoreName so later scheduled runs skip them, and publishedAfter can cut off old entries on a first run. feedFallback flags rows served from the tabs lane when the feed is down.
Pull exact likes and full descriptions
Search rows carry YouTube's rounded figures. Turn on enrichVideoDetails to fetch each video page for likeCount, commentCount, publishedDate and the full description, at one extra request per video. Use it on a shortlist rather than a thousand search rows, and pair likeCount with viewCount for an engagement ratio.
Collect comments and reply threads
Pass a watch URL in startUrls, set scrapeComments true and choose commentsSortBy top or newest. Each comment row gives text, author, authorIsVerified, authorIsCreator, likeCount and replyCount. Set includeReplies true and maxRepliesPerComment to keep threads; replies do not count against maxCommentsPerVideo, which caps top-level comments only.
Build a channel profile row
For any channel URL, includeChannelInfo adds one channel row with channelSubscriberCount, videoCount, totalViewCount, country, joinedDate and links[] to the external sites the channel lists. Pair that with channelContent both to get videos and Shorts, then use maxVideosPerChannel to bound the run while keeping the profile row for a directory or ranking table.
Output fields
| Field | Type | Description |
|---|---|---|
itemType | string/null | Row kind: video, short, channel or comment. |
source | string/null | Where the row came from: search, channel, playlist, video, comments, or rss (a channel or playlist upload feed in channelSource=rss mode). |
input | string/null | The search query or start URL that produced this row. |
feedFallback | boolean/null | Video rows only. true when the RSS feed was unavailable and the row came from the channel Videos tab or playlist lane instead; false for normal feed rows; null for rows of non-RSS inputs. |
position | integer/null | 1-based rank of the item within its input. |
scrapedAt | string/null | ISO timestamp when the row was scraped. |
videoId | string/null | YouTube video identifier. |
url | string/null | Canonical URL of the video or Short. |
title | string/null | Video, Short or channel title. |
description | string/null | Full description (video details only). |
durationText | string/null | Duration as shown by YouTube, e.g. 6:15:56. |
durationSeconds | integer/null | Duration in seconds. |
Show all 40 fields
| Field | Type | Description |
|---|---|---|
viewCount | integer/null | Exact view count when available. |
viewCountText | string/null | YouTube's displayed view count, e.g. 119M views. |
likeCount | integer/null | Exact like count; only with video details enrichment. |
likeCountText | string/null | Rounded like count, e.g. 19M. |
publishedText | string/null | Relative or absolute publish text, e.g. 4 days ago. |
publishedDate | string/null | ISO publish date (yyyy-mm-dd) when exact dates are available. |
channelId | string/null | Channel identifier (UC...). |
channelName | string/null | Channel display name. |
channelHandle | string/null | Channel handle, e.g. @MrBeast. |
channelUrl | string/null | Channel URL. |
channelSubscriberCount | integer/null | Parsed subscriber count, e.g. 522000000. |
channelSubscriberCountText | string/null | Subscriber count as displayed, e.g. 522M subscribers. |
channelVerified | boolean/null | Whether the channel has a verification badge. |
thumbnailUrl | string/null | Thumbnail image URL. |
isLive | boolean/null | Whether the video is a live stream. |
badges | array/null | Visible badges for the video. |
commentCount | integer/null | Parsed comment count; only with video details enrichment. |
commentCountText | string/null | Comment count as displayed, e.g. 2.4M. |
hashtags | array/null | Hashtags extracted from the description. |
isShort | boolean/null | Whether the video is a YouTube Short. |
channelDescription | string/null | Channel description text. |
avatarUrl | string/null | Channel avatar image URL. |
videoCount | integer/null | Number of public videos on the channel. |
totalViewCount | integer/null | Lifetime views across all channel videos. |
country | string/null | Channel country from the About panel. |
joinedDate | string/null | ISO date the channel was created, e.g. 2012-02-19. |
links | array/null | Channel external links as {title, url} objects. |
commentId | string/null | Comment identifier. |
Sample output
From one of our own test runs on 20 Nov 2026, trimmed. Personal data is redacted.
2 rows, first 3 shown
[
{
"itemType": "video",
"source": "rss",
"input": "https://www.youtube.com/@mkbhd",
"position": 1,
"scrapedAt": "2026-11-20T02:45:48.028Z",
"feedFallback": false,
"videoId": "KTDJiVpSDTI",
"url": "https://www.youtube.com/watch?v=KTDJiVpSDTI",
"title": "Can You Trust MKBHD?",
"description": "Can you trust MKBHD videos? Part 2.\n\nPart 1: https://youtu.be/5nLWTVTwOhY\n\nMKBHD Merch: http://shop.MKBHD.com\n\nPlaylist of MKBHD Intro music: https://goo.gl/B3AWV5\n\n~\nhttp://twitter.com/MKBHD\nhttp://instagram.com/MKBHD\nhttp://facebook.co...",
"durationText": null,
"durationSeconds": null,
"viewCount": 3364157,
"viewCountText": null,
"likeCount": 83260,
"likeCountText": null,
"publishedText": "2026-11-17T18:39:57+00:00",
"publishedDate": "2026-11-17",
"channelId": "UCBJycsmduvYEL83R_U4JriQ",
"channelName": "Marques Brownlee",
"channelHandle": null,
"channelUrl": "https://www.youtube.com/channel/UCBJycsmduvYEL83R_U4JriQ",
"channelSubscriberCount": null,
"channelSubscriberCountText": null,
"channelVerified": null,
"thumbnailUrl": "https://i4.ytimg.com/vi/KTDJiVpSDTI/hqdefault.jpg",
"isLive": false,
"badges": null,
"commentCount": null,
"commentCountText": null
},
{
"itemType": "video",
"source": "rss",
"input": "https://www.youtube.com/@mkbhd",
"position": 2,
"scrapedAt": "2026-11-20T02:45:48.098Z",
"feedFallback": false,
"videoId": "hdMszZ2ah5Q",
"url": "https://www.youtube.com/watch?v=hdMszZ2ah5Q",
"title": "Macbook PROLED Review: It's Not About That",
"description": "Incredible laptop. But don't get distracted.\n\nProtect and make your MacBook unique with dbrand at https://dbrand.com/shop/macbook\n\nMKBHD Merch: http://shop.MKBHD.com\n\nPlaylist of MKBHD Intro music: https://goo.gl/B3AWV5\n\nLaptop provided ...",
"durationText": null,
"durationSeconds": null,
"viewCount": 6499110,
"viewCountText": null,
"likeCount": 77616,
"likeCountText": null,
"publishedText": "2026-11-11T18:43:56+00:00",
"publishedDate": "2026-11-11",
"channelId": "UCBJycsmduvYEL83R_U4JriQ",
"channelName": "Marques Brownlee",
"channelHandle": null,
"channelUrl": "https://www.youtube.com/channel/UCBJycsmduvYEL83R_U4JriQ",
"channelSubscriberCount": null,
"channelSubscriberCountText": null,
"channelVerified": null,
"thumbnailUrl": "https://i1.ytimg.com/vi/hdMszZ2ah5Q/hqdefault.jpg",
"isLive": false,
"badges": null,
"commentCount": null,
"commentCountText": null
}
]Input
| Field | Type | Required | Description | Default |
|---|---|---|---|---|
searchQueries | array | No | Keyword searches to run. Each query yields up to 'maxResultsPerQuery' results. | - |
startUrls | array | No | YouTube URLs: videos (/watch?v=, youtu.be), Shorts (/shorts/), channels (@handle, /channel/, /c/, /user/), playlists (?list=), search result pages (/results?search_query=) or hashtags (/hashtag/). | - |
maxResultsPerQuery | integer | No | Maximum rows per search query. | 20 |
maxVideosPerChannel | integer | No | Maximum video/short rows per channel or playlist URL. | 20 |
channelSource | string | No | How channel URLs are scraped. 'tabs' reads the channel Videos/Shorts tabs over YouTube's internal API. 'rss' reads YouTube's public uploads feed (one small request per channel, the 15 newest uploads including Shorts and live streams). Vi... | tabs |
onlyNewVideos | boolean | No | RSS mode: deliver only video ids not delivered by an earlier run of this actor. Delivered ids are kept in a named key-value store in your account (stateStoreName). Leave off for the first run. | false |
stateStoreName | string | No | Named key-value store that holds the delivered ids for onlyNewVideos. 1-63 letters, digits or dashes. | youtube-scraper-state |
publishedAfter | string | No | RSS mode: skip feed entries published before this date or datetime, for example 2026-10-01 or 2026-10-01T00:00:00Z. Leave empty for no cutoff. | - |
channelContent | string | No | Which tab to scrape for channel URLs ('videos', 'shorts' or 'both'). A /@handle/shorts URL always forces Shorts. | videos |
includeChannelInfo | boolean | No | For every channel URL, also deliver one 'channel' row with subscriber counts, About-panel info and links. | true |
searchSortBy | string | No | Sort order for search results. | relevance |
searchUploadDate | string | No | Upload-date filter for search results. | any |
searchIncludeShorts | boolean | No | Also extract Shorts rows from search result shelves. | false |
enrichVideoDetails | boolean | No | Fetch the video page for every result for exact like counts, exact publish date, full description and comment count. One extra request per video. | false |
scrapeComments | boolean | No | Also scrape comments for each video (adds 'comment' rows to the dataset). | false |
maxCommentsPerVideo | integer | No | Maximum TOP-LEVEL comment rows per video. Replies are capped separately by Max Replies Per Comment and do not count toward this limit. | 20 |
commentsSortBy | string | No | Comment order: 'top' (most upvoted) or 'newest'. | top |
includeReplies | boolean | No | Also scrape replies to comments. | false |
maxRepliesPerComment | integer | No | Maximum reply rows per comment thread. Reply rows do not count toward Max Comments Per Video. | 5 |
maxTotalResults | integer | No | Global cap on total rows across all inputs. Hard stop: no rows past this number. | 1000 |
maxConcurrency | integer | No | How many inputs are processed in parallel. Pagination inside one input is sequential. | 5 |
language | string | No | Language code such as en, ms, pt-BR (sent as the interface language hl). | en |
country | string | No | Two-letter country code such as US, MY, GB (sent as the content region gl). | US |
proxyConfiguration | object | No | Optional. Proxies are usually not needed: the scrape works over plain HTTPS from datacenter IPs. Enable Apify Proxy only for high volumes. | {"useApifyProxy":false} |
Example input
{
"searchQueries": [
"python tutorial"
],
"startUrls": [
{
"method": "GET",
"url": "https://www.youtube.com/@MrBeast"
}
]
}Pricing
| Event | What it means | Price | Lowest tier |
|---|---|---|---|
result | One video, short, channel or comment row delivered to the dataset. | $0.001 | $0.001 |
apify-actor-start | Charged when the Actor starts running. Number of events charged depends on Actor memory (one event per GB, minimum one event). | $0.00005 | $0.00005 |
$1.00 for 1,000 results at the listed price.
Prices from the Apify Store on 20 November 2026. Apify shows the final price before each run.
How it compares with other Apify actors
Price per 1,000 results at Apify Store list prices. Every row is a live listing on the Store, read on 20 Nov 2026, so a buyer can check each number at the source.
| Actor | Price per 1,000 results (Free plan) | Higher plans (Gold) | Start fee per run | Platform usage | Users | Rating | Last build |
|---|---|---|---|---|---|---|---|
| YouTube Scraper: Videos, Shorts, Channels & Comments This API | $1.00 | $1.00 | $0.00005 | Included | 10 | No reviews yet | 18 Nov 2026 |
| YouTube Scraper | $4.00 | $2.40 | None | Included | 212,337 | 4.1 (198 reviews) | 18 Nov 2026 |
| Youtube Scraper (Pay Per Result) | $0.5 | $0.485 | None | Included | 56,071 | 3.3 (21 reviews) | 20 Nov 2026 |
| Fast YouTube Channel Scraper | $1.30 | $0.5 | None | Included | 31,595 | 4.6 (44 reviews) | 18 Nov 2026 |
| YouTube Shorts Scraper | $4.00 | $2.40 | None | Included | 73,183 | 4.3 (22 reviews) | 18 Nov 2026 |
| YouTube Comments Scraper | $2.00 | $0.9 | None | Included | 31,890 | 4.9 (42 reviews) | 18 Nov 2026 |
Where YouTube Scraper is stronger
Where YouTube Scraper is stronger
- Can transcribe videos and return subtitles, which we do not offer
- Can add AI descriptions and summaries per video
- Far larger user base and a long review history on the Apify Store
Where this API is stronger
- Plain HTTPS calls to YouTube with no browser and no API key
- Date filtering on channels needs no separate add-on event
- Comments with replies can be collected in the same run as video details
Where Youtube Scraper (Pay Per Result) is stronger
Where Youtube Scraper (Pay Per Result) is stronger
- Much larger user base and a longer review history on the Apify Store
- Accepts YouTube handles and trending requests directly in the input
Where this API is stronger
- One actor covers search, channel videos, Shorts, playlists and comments
- Output includes channel joined date, subscriber counts and hashtags per row
- onlyNewVideos skips what an earlier scheduled run already delivered
Where Fast YouTube Channel Scraper is stronger
Where Fast YouTube Channel Scraper is stronger
- Built only for channels, with sort and oldest post date inputs in the run
- Output includes channel banner URL and channel total views
- Larger user base and more recent runs on the Apify Store
Where this API is stronger
- Channel videos and Shorts, keyword search and comments come from one input set
- An RSS style channel feed keeps new upload tracking cheap on large channel lists
- Plain HTTPS with no browser and no API key
Where YouTube Shorts Scraper is stronger
Where YouTube Shorts Scraper is stronger
- Focused on the Shorts tab, with a dedicated multi channel input
- Output includes channel banner, joined date and location fields
Where this API is stronger
- Shorts, long videos, search and comments come from a single actor
- Date filtering does not trigger a separate add-on charge
- Plain HTTPS with no browser and no API key
Where YouTube Comments Scraper is stronger
Where YouTube Comments Scraper is stronger
- Dedicated comment actor with sorting and oldest comment date inputs
- Output includes replyToCid and creator heart flags per comment
- Far larger user base and a long review history
Where this API is stronger
- Comments can be pulled alongside video details in a single run
- Replies are returned inline, up to the per comment cap you set
- Plain HTTPS with no browser and no API key
These actors return the same YouTube rows, so the decision rests on scope and add-ons. The streamers actors are specialists with transcripts, AI summaries and mature review history, and Apidojo suits very large search jobs. Buyers who want search, channels, Shorts and comments from one actor with no browser and no key will find this one covers that ground.
Prices are Apify Store list prices, checked 20 Nov 2026. Platform usage 'Paid separately' means Apify bills compute on top of the event price. Check each listing before you buy.
How to run it
- Choose your input: searchQueries for keywords, or startUrls for channels, playlists and single videos. Both can run together.
- Set the caps: maxResultsPerQuery for search, maxVideosPerChannel for channel and playlist URLs, and maxTotalResults as a hard stop.
- Add enrichment only where you need it: enrichVideoDetails for exact likes and descriptions, scrapeComments with maxCommentsPerVideo for comments.
- Optionally set language, for example en or ms, and country, for example MY or US, to change the interface and content region.
- Run the actor and export the dataset as JSON, CSV or Excel, or read it through the Apify API.
cURL
curl -X POST "https://api.apify.com/v2/acts/subimpact~youtube-scraper/run-sync-get-dataset-items?token=$APIFY_TOKEN" \
-H "Content-Type: application/json" \
-d '{"searchQueries":["python tutorial"],"startUrls":[{"method":"GET","url":"https://www.youtube.com/@MrBeast"}]}'Python
from apify_client import ApifyClient
client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("subimpact/youtube-scraper").call(run_input={"searchQueries":["python tutorial"],"startUrls":[{"method":"GET","url":"https://www.youtube.com/@MrBeast"}]})
for item in client.dataset(run["default_dataset_id"]).iterate_items():
print(item)JavaScript
import { ApifyClient } from 'apify-client';
const client = new ApifyClient({ token: '<YOUR_APIFY_TOKEN>' });
const run = await client.actor('subimpact/youtube-scraper').call({"searchQueries":["python tutorial"],"startUrls":[{"method":"GET","url":"https://www.youtube.com/@MrBeast"}]});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);MCP
claude mcp add subimpact-data -e APIFY_TOKEN=<YOUR_APIFY_TOKEN> -- npx -y @subimpact/data-mcp
# Then ask your agent to use the youtube tool.n8n
Use the Apify node (Run an Actor).
Actor ID: subimpact/youtube-scraper
Input JSON: paste the example input above.
MCP users: @subimpact/data-mcp exposes these actors as tools, so an agent can call youtube directly.
Limits
- No transcripts or captions, and no video or audio download.
- RSS mode returns at most the 15 newest uploads per channel.
- Exact subscriber counts are not available without login, only YouTube's rounded display figure.
- Duration for a bare video URL can be null when the video is too new to appear in search.
Questions people ask
What does a run cost?
Every delivered video, Short, channel or comment row costs $0.001, which is $1.00 per 1,000 results. There is also a small actor start charge of $0.005. Nothing is charged for rows that fail to be delivered, and compute and memory are included in the price.
Do I need a YouTube API key?
No. The actor calls YouTube's own internal endpoints over plain HTTPS, so there is no key, no OAuth and no headless browser. That keeps the platform usage small, and the same input works from Apify Console, the API, a schedule or an MCP client.
How many videos can I get per channel?
In tabs mode maxVideosPerChannel sets the cap for channel and playlist URLs. In RSS mode you get at most the 15 newest uploads per channel, because that is all the public Atom feed carries. Schedule the monitor more often than a channel posts if it can publish more than 15 videos between runs.
Why are view or subscriber counts rounded?
Those come from YouTube's displayed text, such as 119M views or 522M subscribers. Exact view counts are available where the video details were fetched, and exact like counts appear with enrichVideoDetails. Exact subscriber counts are never exposed to logged out requests.
Can I get transcripts or download the video files?
No. Caption and transcript endpoints are blocked from datacenter addresses, and the player endpoint is blocked for anonymous access, so stream URLs are not provided. The actor returns metadata, comments and channel data only.
How do I watch channels daily without paying twice?
Use channelSource rss with onlyNewVideos true and run on a schedule. Delivered ids are kept per channel in the named key-value store set by stateStoreName, up to 500 ids per channel, and skipped on the next run. A timed-out run can at worst deliver a few ids twice.
Can I scrape one video's comments only?
Yes. Pass the watch URL in startUrls, set scrapeComments to true and leave the search fields empty. You get one video row plus comment rows, with maxCommentsPerVideo capping top-level comments and maxRepliesPerComment capping replies. Sort by top for the most liked or newest for recent discussion.
Need a data source that is not here?
We build and run scrapers for sites that have no public API. Tell us the site, the fields and how often you need them.