YouTube Shorts Scraper (channel Shorts, views, no login)

Get the latest Shorts of any YouTube channel as one flat row per Short: video id, Shorts URL, title, view count, vertical thumbnail and the channel behind it. No API key, no OAuth, no cookies — the actor reads only what YouTube serves to a logged-out visitor on the channel's public Shorts tab.

Run it once for a list, or run it on a schedule in monitor mode and get a webhook or Telegram alert the moment a channel posts a new Short.

What you get

One row per Short: id (video id), url (https://www.youtube.com/shorts/<id>), videoId, title, viewCountText, viewCount, likeCount, publishedAt, durationSeconds, thumbnailUrl, channelId, channelHandle, channelName, channelSubscribers, detailsFetched, rank, scrapedAt.

Rows come back in tab order — newest Short first, rank: 1.

Example output

Field	Example value
id / url	wSh8fAFWEWs / https://www.youtube.com/shorts/wSh8fAFWEWs
title	Survivor vs Beast Games
viewCountText / viewCount	39M views / 39916641 (exact, with includeDetails)
likeCount / publishedAt	921474 / 2026-01-16T11:00:01-08:00 (with includeDetails)
durationSeconds	27
channelHandle / channelId	@MrBeast / UCX6OQ3DkcsbYNE6H8uQQuVA
thumbnailUrl	https://i.ytimg.com/vi/wSh8fAFWEWs/oardefault.jpg?...

Use cases

• Track what a set of creators is posting in Shorts and how each Short performs, without opening YouTube.
• Build a content-research list for one niche: 30-48 recent Shorts per channel with titles and view counts.
• Watch competitors on a schedule and get alerted on every new Short (monitor mode keys on the video id).

How it works

1. Each channel's public Shorts tab (https://www.youtube.com/@handle/shorts) is opened as a logged-out visitor. The page embeds the whole grid, so one request returns every Short the tab renders — typically 30 to 48, newest first.
2. With includeDetails on, each Short's own page (https://www.youtube.com/shorts/<id>) is opened for the exact view count, like count, publish date and duration (one extra request per Short).
3. A bot-check or consent wall is retried on a stronger proxy tier before the run gives up, and a walled run is reported as blocked — an empty dataset never hides a wall.

Input

Field	Meaning	Default
channels	Channels to read, one per line: @MrBeast, a channel URL, or a UC... id	["@MrBeast"]
maxShortsPerChannel	Stop after this many Shorts per channel (1-50)	48
includeDetails	Open each Short for exact views, likes, publish date, duration	false
tier	Proxy tier: auto climbs none → datacenter → residential on a wall	auto
proxyCountry	Exit country for the proxy	US
maxItems	Stop after this many rows in total	100
mode	scrape or monitor (only new/changed since the last run)	scrape
monitorStateId, webhookUrl, telegramBotToken, telegramChatId	Monitor state key and alert targets	empty

Pricing

Event	Price
result	$0.001 per Short ($1 per 1,000)
monitor-check	$0.005 per monitor run
change	$0.001 per new or changed Short

You are charged only for rows actually pushed. A run that finds nothing pushes nothing and charges no result events. Proxy traffic is billed by Apify on top, and only when a wall forces the run above the free none tier.

Monitor mode and includeDetails. A change event is billed per row whose view count, like count or title moved since the last run. With the default (includeDetails: false) the view count is YouTube's own rounded figure (39M views), which stands still for hours or days — so a monitor run normally bills only genuinely new Shorts. With includeDetails: true the exact viewCount is read instead, and on an active channel that number is different on every run: expect a change event for essentially every Short you watch, on every check, plus one extra page request per Short. Watch for new Shorts with the default, and turn details on only when you want the exact curve and have budgeted for it.

Limits

• One page per channel. YouTube's robots.txt disallows /youtubei/, the endpoint that loads "more Shorts", so this actor takes only what the first page of the Shorts tab renders: 30-48 Shorts, newest first. It is not an archive tool for a channel's entire Shorts history.
• No Shorts search. robots.txt also disallows /results, so there is no keyword search here — the input is a list of channels.
• Without includeDetails the view count is the rounded number YouTube prints on the tab (39M views → 39000000); likeCount, publishedAt and durationSeconds are empty. Turn includeDetails on for exact numbers.
• A channel with no Shorts, or a handle that does not exist, returns no rows and costs no result events.
• Counts are what YouTube shows a logged-out visitor and can lag the creator's own analytics.

What is NOT collected

No comments, no commenter names, no subscriber lists, no viewer data, no personal profiles — nothing about private individuals. Rows describe a public video and the channel that published it, which is a creator/business entity.

FAQ

Does it need a YouTube account, API key or cookies? No. There is no credential field at all; every request is made as a logged-out visitor.

Why do I get 48 Shorts and not 200? Because loading Shorts 49+ requires an endpoint YouTube's robots.txt disallows. We do not build against an explicit Disallow; see the Limits section.

What happens when a run comes back empty? No rows are pushed and no result events are charged. The RUN_SUMMARY record in the run's key-value store carries emptyReason, which separates "this channel has no Shorts" from "YouTube answered with a bot check" — the second is reported as blocked, never as no_matches.

Changelog

• 0.1: initial release — channel Shorts tab, optional exact per-Short stats, proxy-tier ladder on the bot wall, monitor mode for new Shorts.

---

If this actor saved you time, a short review on its Store page genuinely helps other people find it. Found a bug or need a field that is missing? Open a ticket on the Issues tab.