YouTube Channel Scraper - Videos & Shorts

YouTube channel scraper - videos, Shorts, live streams, playlists & posts

One actor for a whole YouTube channel. Give it a handle, a URL, or a channel ID and get every video, Short, live stream, playlist and community post, plus the channel's own metadata.

$1.00 per 1,000 results. $0.005 per run. No browser, no proxy setup.

Works on any public channel. Metadata comes back in English so counts and dates parse reliably; titles and descriptions are the creator's own text.

Why this scraper?

• 📅 Exact publish timestamps: most channel scrapers hand you "3 days ago" and let you guess. Turn on enrichment and every video carries the real upload time.
• 🔢 Exact view counts: not "2.3M views" rounded to two digits, the real integer.
• 🎬 Shorts get the same treatment: YouTube's Shorts listings carry a title and a rounded view count and nothing else. We fill in duration, exact date, category, description and tags anyway.
• 📚 The whole catalog, not the first page: thousands of videos from a single channel, paged to the end. There is no page cap on an unfiltered run.
• 🧾 Honest nulls: when a view count is hidden or a field genuinely isn't published, you get null, never a fabricated number. Failed enrichment is labelled in enrichmentError rather than silently blank.
• 💸 Bad input doesn't bill you: a channel that fails to resolve is skipped with a warning, and you're charged only for rows that land in your dataset.

What you can pull

Content	Input	Notes
Long-form videos	maxVideos	Newest first, no depth limit
Shorts	maxShorts	Same enrichment as regular videos
Live streams	maxLiveStreams	Past, in progress and scheduled, flagged with wasLiveStream, isLive and isUpcoming
Playlists	maxPlaylists	Title, video count, first video, thumbnail
Community posts	maxCommunityPosts	Text, attached images, poll questions and options
Channel profile	always	Subscribers, total views, country, join date, links, monetization signals

Set any limit to 0 to skip that content type and keep the run cheap.

Use cases

• Creator analytics: track upload cadence, median views and engagement across a roster of channels on a schedule.
• Competitive intelligence: watch what competitors publish, how often, and which formats they lean into.
• AI and LLM pipelines: pull titles, descriptions, tags and categories as structured training or retrieval data.
• Sponsorship research: size a channel properly before you pay for placement, using real view distributions instead of a headline subscriber count.
• Trend monitoring: sweep a keyword for channels, then track what the whole cohort ships.

Input

Paste channels one per line. Handles, full URLs and UC… IDs all work, and you can mix them.

{
  "channels": ["@MrBeast", "https://www.youtube.com/@mkbhd", "UCXuqSBlHAE6Xw-yeJA0Tunw"],
  "maxVideos": 50,
  "maxShorts": 20,
  "enrichVideos": true,
  "includeDerivedAnalytics": true
}

No channel list? Set searchKeyword instead and the actor finds channels for you.

Output

One row per item. Every row carries a type field so you can filter, and video rows also carry videoKind (regular, short or live).

A video row with enrichment on. Note viewCount: 52 beside viewCountText: "51 views": that isn't a bug. The text is what YouTube rendered on the channel page, which is rounded and cached; the number is the exact live count from enrichment.

{
  "type": "video",
  "videoKind": "regular",
  "videoId": "fUymQq5llCo",
  "title": "Build a Delivery Competitor Monitor",
  "url": "https://www.youtube.com/watch?v=fUymQq5llCo",
  "viewCount": 52,
  "viewCountText": "51 views",
  "publishedAt": "2026-07-24T17:00:12.000Z",
  "publishedText": "15 hours ago",
  "publishedAtIsApproximate": false,
  "durationSeconds": 219,
  "isLive": false,
  "wasLiveStream": false,
  "isMembersOnly": false,
  "description": "See how you can automatically track every location...",
  "tags": ["web scraping platform", "web automation", "scrapers"],
  "category": "Science & Technology",
  "exactPublishedAt": "2026-07-24T17:00:12.000Z",
  "exactViewCount": 52,
  "channelId": "UCTgwcoeGGKmZ3zzCXN2qo_A",
  "channelHandle": "@Apify"
}

A channel row:

{
  "type": "channel",
  "channelId": "UCXuqSBlHAE6Xw-yeJA0Tunw",
  "handle": "@LinusTechTips",
  "name": "Linus Tech Tips",
  "country": "Canada",
  "joinedAt": "2008-11-24T00:00:00.000Z",
  "subscriberCount": 16900000,
  "totalViews": 9686730358,
  "totalVideos": 7800,
  "totalVideosIsApproximate": true,
  "verified": true,
  "socialLinks": { "twitter": "https://twitter.com/LinusTech" },
  "monetization": { "hasMembership": true, "hasMerchShelf": false }
}

Playlist and community-post rows follow the same pattern. Filter on type to split them out.

How much does it cost?

Two charges, that's it:

• $1.00 per 1,000 results. Every row costs the same, whether it's a plain video row or a fully enriched one.
• $0.005 per run.

On the Apify free plan ($5 of credit) that's about 5,000 results a month. On Starter ($29), about 29,000.

Enrichment and like counts add requests, not charges. A 50-video channel costs about $0.055 either way.

Limitations

Worth knowing before you run it:

• Date filters always use exact dates. Channel listings only say "2 months ago", which is too coarse to filter on, so setting publishedAfter or publishedBefore switches enrichment on automatically and every returned row is checked against its real publish timestamp.
• Very narrow windows on very busy channels are slow. Channel listings date a video only to the nearest week, month or year depending on its age, so the actor checks individual videos whenever that granularity is coarser than your window. A one-day window on a channel that posts dozens of times a day is the worst case and can take a few hundred requests. Widening the window, or raising the videos limit, returns far more rows for the same work. Depth is not the problem: a window several years back is often quicker than one a few weeks back, because older videos sit in wide year buckets that are fast to rule out wholesale.
• Historical windows on Shorts are limited. Shorts listings carry no date at all, so nothing can be ruled out without checking it. The run does as much as its search budget allows and then reports a partial result. For older Shorts, narrow the range and run one channel at a time.
• A window deep in a large catalog may not be fully reached. The run works backwards from the newest upload within a search budget; if it runs out before covering the whole range it says so in the status rather than pretending it found everything. Narrow the range or run one channel at a time.
• totalVideos is rounded on big channels. YouTube reports "7.8K videos" rather than an exact figure, so totalVideosIsApproximate tells you when the number came from a rounded label.
• A small share of videos fail enrichment. YouTube rate-limits some requests. Those rows come back with enrichmentError set and their listing fields still populated. The one exception is a date-filtered run: any video whose enrichment fails has no trustworthy date to filter on, so it is retried once and then skipped rather than guessed at, and the run status tells you how many were skipped.
• Like counts are opt-in and slow. YouTube exposes likes only on a payload roughly 70 times larger than the one carrying everything else, so includeLikeCount is a separate switch.
• Members-only and private videos appear in listings but can't be enriched. They're marked, not dropped.
• Scheduled premieres have no publish date. They come back with isUpcoming: true and a null publishedAt, and they're excluded from date filters, because they haven't aired yet.
• Derived analytics describe the videos you collected, not the channel's whole history. Ask for more videos and the median moves.
• Poll vote counts aren't available. YouTube doesn't publish them, so poll options come back without tallies.
• Comment text and transcripts aren't here. This actor covers the channel and its catalog. Those belong to a video scraper.

FAQ

Do I need a YouTube API key? No. There's no key, no login and no quota to manage. Paste channels and run.

Do I need to configure a proxy? No. Datacenter proxy is used by default and is enough for YouTube. If an address does get blocked, the actor retries on a residential one automatically.

Can I run this on a schedule? Yes. Apify Schedules can run it hourly, daily or weekly, and each run appends to a dataset you can pull over the API.

How do I get only recent videos? Set publishedAfter to a date like 2026-01-01 or a relative value like 30 days. Both bounds include the named day, and enrichment turns on automatically so the cutoff is exact.

What happens if a channel doesn't exist? It's skipped with a warning, the rest of your list still runs, and you aren't charged for it. The run status message tells you how many channels failed.

Your feedback Found a bug or need a field that isn't here? Open an issue at console.apify.com/actors/SERgpqPKFdbZrAOmq/issues and it goes straight to a human.