YouTube Channel Shorts Scraper

Collects Shorts that YouTube publicly lists on one channel’s Shorts tab. It uses the page’s public ytInitialData payload and the same public browse continuations used by the tab, rather than relying on page scrolling or hidden/account-only data.

Input

Field	Required	Description
channelUrl	Yes	A public @handle, UC… channel ID, or canonical https://www.youtube.com/@handle / https://www.youtube.com/channel/UC… URL.
maxItems	No	Maximum public Shorts to return, from 1 to 500 (default 50).
proxyConfiguration	No	Apify proxy configuration. Residential proxies are useful if YouTube rate-limits the run.

The actor intentionally does not expose a sort selector: the public Shorts tab does not provide a stable, documented sort control, so results preserve YouTube’s returned order.

Example:

{
  "channelUrl": "https://www.youtube.com/@NASA",
  "maxItems": 25,
  "proxyConfiguration": {
    "useApifyProxy": true,
    "apifyProxyGroups": ["RESIDENTIAL"]
  }
}

Output

Every record is a source-backed public Shorts lockup. Empty, null, non-finite and unavailable fields are omitted rather than manufactured.

Field	Meaning
position	Returned order on the public Shorts tab.
inputChannelUrl, sourceUrl, sourceTab	Input and public source provenance.
channelName, channelId, channelHandle, channelUrl	Public channel metadata when exposed by YouTube.
videoId, title, url, thumbnail	Short identity and public media links.
viewCountText, viewCountApprox	Views as displayed and a best-effort numeric parse when source text supports it.
publishedText, publishedDate, duration	Only emitted when the Shorts payload exposes them. publishedDate is emitted only for an ISO date supplied in the source.
isShort	Always true, backed by YouTube rendering the item in its Shorts tab.
scrapedAt	ISO 8601 collection timestamp.

The actor fails the run when no public Shorts are returned, which prevents an apparently successful empty dataset from being mistaken for valid output.

Operational notes

• The installed Chrome build supplies its own matching user agent; bounded retries use fresh Crawlee sessions without browser/UA mismatches.
• OUTPUT_SUMMARY records the final source status, output count, and proxy use separately from the Shorts dataset.
• Invalid proxy configuration fails fast. The actor does not silently fall back to a direct connection.
• Only public pages and public browse responses are used. It never attempts to access private, age-gated, account-only, or authentication-gated data.

Example output record:

{
  "recordType": "youtube_channel_short",
  "position": 1,
  "videoId": "dQw4w9WgXcQ",
  "title": "A public Shorts title",
  "url": "https://www.youtube.com/shorts/dQw4w9WgXcQ",
  "isShort": true,
  "sourceTab": "shorts",
  "inputChannelUrl": "https://www.youtube.com/@NASA",
  "sourceUrl": "https://www.youtube.com/@NASA/shorts",
  "scrapedAt": "2026-08-16T10:00:00.000Z"
}