YouTube Channel Scraper & Tracker

Scrape YouTube channels without a login, an API key or Google quotas: exact total views and video count, subscribers, joined date, country, description and links, plus each channel's videos, Shorts, live streams and most popular videos: title, views, age, length, live and upcoming marks. Paste handles (@mkbhd), channel links or channel ids, up to 1,000 at a time.

It is also a YouTube channel tracker and monitor. Schedule it with Only changes and each run returns one row per channel with what it gained since the last run (subscribers, total views, videos, and over 7 and 30 days), and only what moved: new videos, removed videos, changed titles, streams that went live and view jumps, posted to Slack, Discord, Telegram or Microsoft Teams if you like. It also runs as a live API and MCP server for AI agents.

One price, no extras: $1.00 per 1,000 results, with platform usage included and no row cap on Apify's free plan. (Apify's Store page lists platform usage too, because the live API below adds it; ordinary runs never pay it.)

Why this scraper

• It tracks channels, not just lists them. Most YouTube channel scrapers return the same rows every run, and you work out what changed. This one keeps a history per channel, with a reading a day, and returns only what is new.
• A title test isn't a title change. YouTube lets creators test several titles at once, so different viewers see different titles for the same video. A title this tracker has seen before is not reported again, and a real change comes with every title the video has shown.
• Removed means removed. A video that leaves a channel's list is checked with YouTube itself: deletedOrPrivate when YouTube says it is gone, hiddenFromChannel when it still exists off the channel's lists.
• Exact where YouTube is exact. The channel's total views and video count are exact numbers, and they move every day, so totalViewsGained works for channels whose rounded subscriber count barely changes.
• Better than a one-off look. Median views per kind (videos, Shorts, streams), upload pace, days since the last upload, views a day, and how each video did against the channel's usual one (viewsVsMedian: 3.2 is 3.2 times its median).
• No login, no API key, no quota. The Actor reads the public channel pages and video lists anyone can open.
• Honest numbers. Fields carry what the pages show. What a list doesn't show (see What it can't do) is left empty, never guessed.

	Typical YouTube channel scrapers	This Actor
Subscribers, total views, video count, joined date, country, links	Yes	Yes
Videos, Shorts, live streams	Yes	Yes, plus the most popular videos
New and removed videos, title changes since the last run	You compare runs yourself (a few small monitors do new videos)	Built in (Only changes)
Subscriber and view growth, over 7 and 30 days	Rare	Built in, with a reading a day
Title tests told apart from real changes	Rare	Yes
Alerts to Slack, Discord, Telegram, Teams	Rare	Yes
Median views, upload pace, views against the channel's usual	Some	Yes, on every channel row
MCP server for AI agents	Rare	Its own (4 tools)
Exact views and exact dates per video, likes, every video a channel ever made	Those that open every video	No: see What it can't do

Use cases

• Competitor monitoring: a daily schedule on your rivals' channels that posts each new video, title change and view jump to Slack.
• YouTube subscriber tracking: subscribers, total views and video count per day for your channels and your competitors', with 7-day and 30-day gains.
• Influencer research: compare hundreds of channels by median views, upload pace, country and last upload before you reach out. Channel rows only gives one row per channel.
• Content research: which of a channel's videos beat its own median, and which titles it tested.
• News and brand monitoring: know when a channel posts, goes live, or quietly removes a video.
• AI agents: ask an assistant "compare @mkbhd and @LinusTechTips by median views and upload pace" through the MCP server.

How to use it in 4 steps

1. Add Channels: @mkbhd, https://www.youtube.com/@mkbhd, https://www.youtube.com/channel/UCBJycsmduvYEL83R_U4JriQ or UCBJycsmduvYEL83R_U4JriQ.
2. Set how many Videos, Shorts, Live streams and Popular videos to read per channel (100 at most each). Set all four to 0 for channel figures alone, or switch on Channel rows only for one row per channel with its medians and pace.
3. Set Max results and any filters, and run it.
4. To track changes, switch on Only changes since the last run, add an alert if you want one, and schedule it.

What you get

One row per channel (type: "channel") and one per video (type: "video"). A real channel row, shortened:

{
    "type": "channel",
    "input": "@mkbhd",
    "channelId": "UCBJycsmduvYEL83R_U4JriQ",
    "handle": "@mkbhd",
    "name": "Marques Brownlee",
    "channelUrl": "https://www.youtube.com/@mkbhd",
    "subscribers": 21400000,
    "subscribersText": "21.4M subscribers",
    "totalViews": 5856382564,
    "videoCount": 1863,
    "joinedDate": "2008-03-21",
    "country": "United States",
    "isVerified": true,
    "longVideoCount": 1729,
    "shortsCount": 125,
    "streamCount": 6,
    "medianViews": 5150000,
    "medianShortsViews": 5050000,
    "medianStreamViews": 945000,
    "lastUploadAbout": "2026-10-29",
    "uploadsLast30Days": 6,
    "uploadsPerMonth": 8.3,
    "uploadIntervalDays": 3.7
}

A channel row also has the description, keywords, links, avatar, banner and the tabs the channel shows. A real video row:

{
    "type": "video",
    "channelId": "UCBJycsmduvYEL83R_U4JriQ",
    "handle": "@mkbhd",
    "channelName": "Marques Brownlee",
    "videoId": "2reQut8WNGc",
    "url": "https://www.youtube.com/watch?v=2reQut8WNGc",
    "title": "Macbook PROLED XDR Impressions: They Said They Wouldn't!",
    "videoType": "video",
    "views": 10000000,
    "viewsShown": "10M",
    "publishedText": "3 days ago",
    "publishedAbout": "2026-10-29",
    "ageDays": 3.0,
    "durationSeconds": 654,
    "durationText": "10:54",
    "position": 1,
    "lists": ["videos"],
    "viewsPerDay": 3333333,
    "viewsVsMedian": 1.94
}

Live streams add isLiveNow, watchingNow, isUpcoming and scheduledFor (in UTC); a premiere adds isPremiere, and a scheduled stream whose time passed days ago adds scheduleLapsed instead of isUpcoming (T-Series still lists streams scheduled for 2018). Popular videos add popularRank. A video made with other channels adds collaborators, which names all of them, also those YouTube's list hides behind "and 2 more". A video on two lists (the newest and the most popular) is one row, naming both in lists. A name YouTube has no channel for gets a free notFound row.

Titles: title is what the channel wrote. YouTube shows English readers a translation when the creator added one (Felipe Neto's Portuguese titles, HikakinTV's Japanese ones), so each list is read twice, and that translation comes in titleEnglish. Switch off The channel's own titles to read each list once, in about half the time, with the titles English readers see.

Numbers: subscribers and views are as YouTube rounds them for everyone (21.4M, 10M), to 3 digits at most. totalViews and videoCount are exact. YouTube's lists of a channel's videos stop at about 20,000, so longVideoCount stops there too, while videoCount doesn't. Medians leave out streams that never ran (0 views, no length). Dates: lists word a video's age ('3 days ago', '2 years ago'), so publishedAbout is about, and only as fine as those words.

Regions: YouTube shows each country what it may watch there, and the Actor reads from Apify's servers in the United States, or wherever your proxy is. When a list hides videos (deleted, private or blocked in that region), the channel row names it in hiddenOnLists, and the run says so. SET India's list showed 25 videos from Apify's servers and 56 from a home connection, so for a channel like that, use Proxy in its own country if you have one there.

With Only changes, a channel row adds subscribersGained, totalViewsGained, videosAdded, viewsPerDay, daysSinceLastRun, subscribersGained7d, totalViewsGained7d, subscribersGained30d, totalViewsGained30d, previousName and previousHandle when they changed, and history (the last 30 daily readings). Video rows add change:

change	meaning
newVideo	a video, Short or stream posted since the last run
removed	it left the channel: removalKind is deletedOrPrivate (YouTube says it is gone) or hiddenFromChannel (it still exists, off the channel's lists)
titleChanged	a title this tracker has never seen for the video, with previousTitle and titlesSeen
wentLive	a stream or premiere that went live; isLiveNow says whether it still is
viewsJump	views rose past your View jump (%) since the last report, with previousViews, viewsGained and viewsGainedPercent (ask for it in Changes to return)
firstRun	the first run's baseline: the videos there now

The dataset has views for Overview, Videos, Channels, Growth, Changes and Live (a view picks columns: every view lists every row, so filter on type or change if you need to).

How to use it

Channels

A channel is a handle (@mkbhd, or mkbhd), a link (youtube.com/@mkbhd, youtube.com/channel/UC..., youtube.com/user/..., youtube.com/c/..., or a bare youtube.com/name) or a channel id (UC...). Two names of one channel are read once. An old /c/ link that no longer answers is tried as a handle, and the run says so.

Lists and limits

Lists read newest first, the popular list by YouTube's own ranking. Max results (1,000 by default) is shared between channels, in rounds, so a big channel doesn't use up a small one's share; on one channel it keeps some of each list. Rows are never cut at random: what doesn't fit is counted in the run's status message. Channel rows only returns one row per channel, without a row per video.

Filters

Posted after (a date or 7 days), Minimum views, Only titles containing and Exclude words apply to video rows. Words are looked for in the channel's own title and in titleEnglish. Dates are about, as lists word them, and Shorts show none, so Posted after doesn't filter Shorts. Minimum views leaves out streams live now, which show no views.

Tracking changes on a schedule (YouTube channel tracker)

1. Switch on Only changes since the last run. Optionally pick the Changes to return.
2. Run it once: the first run returns each channel's row and its videos as firstRun, and saves them as the baseline.
3. Schedule it (Apify Console, Schedules). Each later run returns a row per channel with its gains, and only the changes.

How it decides:

• New: lists run newest first, so a video above the first known one is new, unless its age says it came before the last run. Live and upcoming videos are new wherever they sit. A video below the first known one moved into view and is recorded without a row. Videos per channel sets how many newest videos a run watches. A stream first seen while live is one wentLive row with isNew: true. A new video under Minimum views is reported on the run that it passes them.
• Removed: a watched video that left its list while older ones stayed is checked with YouTube's public video check. If YouTube says it is gone, it is reported at once. If it still exists, or YouTube doesn't answer, it is reported when it has been missing two runs in a row, because a list can leave a video out for one fetch.
• Changed title: titles are kept per video, so a title YouTube is testing doesn't come back as a change each time it shows, and a new translation isn't a new title. The tracker also keeps one anonymous YouTube visitor per history, as a returning browser does, because different visitors are shown different titles.
• View jumps: per-video views are rounded, so a jump is measured from the last report, and it counts from 1,000 views: 500 to 5,000 is a jump, 200 to 900 isn't. Use it for watched videos; a fresh video always jumps, so it isn't on by default.
• Growth: totalViewsGained and videosAdded are exact. subscribersGained moves in YouTube's rounded steps: a channel at 21.4M shows no change until it reaches 21.5M. YouTube also updates a channel's total views in steps, not by the minute: 12 channels, MrBeast's among them, showed no gain across 3 hours, so runs close together can show none.
• Histories are kept per channel and filters (two trackers with different filters don't mix), in the key-value store youtube-tracker-history, or one you name in History name.
• Only what a run returns moves the history: if Max results or your spending limit stops a run, the changes it didn't return come with the next run.

Alerts in Slack, Discord, Telegram or Microsoft Teams

Paste an incoming-webhook URL in Alert webhook URL (Slack, Discord, Microsoft Teams, or any HTTPS endpoint that takes JSON, such as Zapier, Make or n8n), or a Telegram bot token and Telegram chat id. Each switches on Only changes. The first run posts one short "alerts are set up" message; each later run posts its changes (new videos by default, or those you pick in Alert on), up to 50 per run. A run that stops early (aborted, or failing partway) still posts the changes it stored. A restarted run doesn't post a message twice, and a failed post doesn't fail a run whose data loaded: its status says why.

Run summary for pipelines

Each run saves a RUN_SUMMARY record in its key-value store: per channel, its rows and what each list held; plus the changes, what didn't load and why, and the filters used.

YouTube channel API: use it from code or from an AI assistant

Start a run and get the rows in one call (replace <YOUR_TOKEN>; a call that waits gives up after 5 minutes, so start bigger runs without waiting and read the dataset afterwards):

curl -X POST "https://api.apify.com/v2/acts/akhil_rajesh~youtube-channel-scraper/run-sync-get-dataset-items" \
  -H "Authorization: Bearer <YOUR_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{"channels": ["@mkbhd", "@Fireship"], "maxVideos": 30}'

Or from Python with apify-client:

from apify_client import ApifyClient

client = ApifyClient("<YOUR_TOKEN>")
run = client.actor("akhil_rajesh/youtube-channel-scraper").call(run_input={
    "channels": ["@mkbhd", "@Fireship"], "maxVideos": 30, "maxShorts": 10, "postedAfter": "30 days",
})
for row in client.dataset(run.default_dataset_id).iterate_items():
    print(row["type"], row.get("name") or row.get("title"))

For AI agents: MCP server and live API

The Actor also runs as a live endpoint (Apify Standby): an MCP server that agents call directly, and plain HTTP endpoints. Once a Standby run is up, a request is answered in seconds; the first request starts one. Authenticate with your Apify API token in the Authorization: Bearer <token> header.

MCP server (Streamable HTTP): https://akhil-rajesh--youtube-channel-scraper.apify.actor/mcp

{"mcpServers": {"youtube-channels": {"type": "http", "url": "https://akhil-rajesh--youtube-channel-scraper.apify.actor/mcp",
  "headers": {"Authorization": "Bearer <YOUR_APIFY_TOKEN>"}}}}

In Claude Code:

$claude mcp add --transport http youtube-channels https://akhil-rajesh--youtube-channel-scraper.apify.actor/mcp --header "Authorization: Bearer <YOUR_APIFY_TOKEN>"

A client that can't send headers can use https://akhil-rajesh--youtube-channel-scraper.apify.actor/mcp?token=<YOUR_APIFY_TOKEN>; keep that URL private, as it carries your token.

Tool	What it returns
get_channel_info	One row per channel: subscribers, exact total views and video count, joined date, country, description, links
get_channel_videos	A channel's videos, Shorts, streams and popular videos, with filters
compare_channels	One row per channel, biggest first, with median views, upload pace and days since the last upload
run_actor	The Actor's whole input in one call

HTTP endpoints, such as GET https://akhil-rajesh--youtube-channel-scraper.apify.actor/videos?channels=@mkbhd|@Fireship&videos=30 (separate several values with |). The Actor's Endpoints tab in Apify Console lists them too, and lets you try them.

Endpoint	Parameters
GET /channel	channels
GET /videos	channels, videos, shorts, streams, popular, max, after, keywords, exclude, minViews, originalTitles
GET /compare	channels
POST /run	The Actor's whole input as JSON

A tool call reads up to 5 channels and 200 rows, an HTTP request 20 channels and 2,000 rows. A request stops at its time and size limits and names what it left in notRead. Tracking and alerts belong to scheduled runs: a request is answered on its own. On the live endpoint, platform usage is added to the price of the rows while the run serves you, and a request keeps reading until its limit even if the caller stops waiting.

How much does it cost?

You pay $1.00 per 1,000 results, less on paid Apify plans, plus a start fee of $0.00005 per run. A result is a channel row or a video row. Platform usage is included in runs of the Actor: there is no compute charge on top. (The Store page's pricing box mentions platform usage because of the live endpoint, which adds it; ordinary runs never do.) Names that weren't found are free.

Apify plan	Per 1,000 results
Free	$1.00
Starter	$0.90
Scale	$0.80
Business and above	$0.70

Run	Results	Cost (Free plan)
One channel and its last 30 videos	31	about $0.03
100 channels with Channel rows only	100	about $0.10
100 channels and 10 videos each	1,100	about $1.10
One channel's four lists, full	up to 401	up to $0.40
33 large channels, all four lists	6,669	about $6.67
10 channels tracked daily, about one new video each	about 20 a day	about $0.60 a month

Apify's free plan ($5 of monthly credit) covers about 5,000 results. To cap spending, lower Max results or set a maximum cost per run in the run options: the run stops at the limit and says so.

What it can't do

• Exact views, likes, comments and video descriptions: these need each video's own page, and those pages ask visitors from cloud servers to sign in. This Actor doesn't get around that. Views are the list's rounded figure.
• More than the newest 100 per list: that is what a channel's list shows without YouTube's API. Together, the four lists can hold more than 100 videos a channel. Older videos can't be read, so this isn't a full archive of a channel.
• Dates on Shorts: YouTube's Shorts list shows views, but no dates or lengths.
• Search: finding channels by keyword isn't offered: YouTube's rules for automated visitors exclude its search pages.
• Emails: YouTube keeps channels' business emails behind a sign-in. A channel's description comes as its owner wrote it.
• Subscriber numbers more exact than YouTube's: nobody outside YouTube has them.

FAQ

Do I need a YouTube API key or a Google account? No.

How often can I run it? Daily suits growth tracking. For a channel that posts often, hourly works too. A run takes about 2 seconds per page: one page per channel, and two per list you ask for (one with The channel's own titles off).

Is it legal to scrape YouTube channels? This Actor reads public pages only and collects no personal data beyond what a channel publishes about itself. Check your own use against the laws that apply to you and YouTube's terms.

Is this an official YouTube product? No. This Actor isn't affiliated with or endorsed by YouTube or Google. YouTube is a trademark of Google LLC.