YouTube Playlist Scraper: a hundred videos a query

A list you know, or a list you are looking for

Seven inputs, two of which cost money, and they cost the same. Both carry a hundred free videos, so the choice between naming a playlist and searching for one is about what you want back rather than about price.

• A hundred free videos, three times any sibling

  startUrls

  $0.05per playlist

  100 videos in

  Playlist URLs. The direct route, and the one to use when the list is already chosen.

  https://www.youtube.com/playlist?list=PL…

• keywords

  $0.05per search

  100 videos in

  Terms to find playlists with. Same fee and same allowance as naming one outright.

  ["python tutorial"]

• gl / hl / sort

  us / en / rthe defaults

  Search only

  Region, language and ordering for a keyword search. Free to set, and the listing gives r as the relevance default.

  gl: "DE", hl: "de"

• maxItems

  ÷ $0.0005past the free 100

  Run-wide

  Total videos across the run. On a course playlist of eighty this never binds. On an auto-generated mix it does.

  500

Because the allowance is a hundred either way, most real playlists cost exactly one query fee and nothing else. An album, a course module or a conference track will usually come back complete for a nickel.

A course, a shelf of them, or a search for more

Three runs. The first is the common case and usually carries no item charges at all. The other two are what you do once you have decided the format is worth studying.

{
  "startUrls": ["https://www.youtube.com/playlist?list=PL…"],
  "maxItems": 100
}

{
  "id": "lL2ZwXj1tXM",
  "title": "Three Days Grace - Never Too Late",
  "url": "https://www.youtube.com/watch?v=lL2ZwXj1tXM",
  "duration": 211,
  "views": 272272914,
  "status": "OK",
  "isLive": false,
  "isPrivate": false
}

{
  "startUrls": [
    "https://www.youtube.com/playlist?list=PL…",
    "https://www.youtube.com/playlist?list=PL…"
  ],
  "maxItems": 500
}

{
  "channel": {
    "id": "UCK-VRwguZP3cokOAxI0F1HA",
    "name": "ThreeDaysGraceVEVO",
    "url": "https://www.youtube.com/channel/UCK-VRwguZP3cokOAxI0F1HA"
  }
}

{
  "keywords": ["python tutorial"],
  "gl": "us",
  "hl": "en",
  "maxItems": 200
}

{
  "description": "Three Days Grace's official music video...",
  "keywords": ["three days grace", "rock", "music", "official video"]
}

Ten fields promised, two not

A generous availability table by this family’s standards. The engagement figures are guaranteed here, where the Channel Scraper leaves its subscriber count optional, and only the free-text fields are conditional.

• 3

  Always present

  Identity

  • id, the video ID
  • title
  • url, the watch link

• 3

  Always present

  Typed as numbers

  • views
  • likes
  • duration, in seconds

• 4

  Always present

  Context

  • channel.id, channel.name
  • thumbnails, array of sizes
  • isLive, a boolean

• 2

  Sometimes present

  The free text

  • description
  • keywords, array of tags

duration being a guaranteed number is what makes this actor useful for course auditing. Summing it across a playlist gives the real length of a curriculum, which is the question anyone evaluating training content actually asks first.

Worth knowing before the first run

Two of these come from the listing’s availability table, two from its pricing section, and one from a discrepancy inside its own output sample.

• 01

  A hundred videos come free with every query

  Three times what the Channel Videos actor includes and five times the Comments actor. On lists of a hundred or fewer the item rate never applies at all.

• 02

  Searching costs the same as naming

  Both the playlist query and the keyword search are $0.05 with a hundred videos included. There is no cheaper door here, which is not true of the two channel scrapers.

• 03

  The output sample repeats one number twice

  It gives views and likes as the same value, 272272914, which cannot be right for a real video. Trust the availability table, where both are typed as numbers, rather than that pair of figures.

• 04

  There is no playlist position field

  The listing mentions "position in playlist" in prose, but no such field appears in its output sample or its availability table. Order arrives as dataset order, not as a value you can sort on later.

• 05

  Private entries still cost

  isPrivate appears in the sample, so a playlist containing videos you cannot watch will still return and bill rows for them. Filter after the fact, not before.

Free accounts run in demo mode: 5 runs a month, 10 items each.

Where the hundred runs out

These are the listing’s five published scenarios plus a search, all reproducing from its two charge events. Read them together and the shape is clear: below a hundred videos a playlist is a flat nickel, and above it the item rate takes over slowly.

Rows four and five are the comparison worth making. Five hundred videos cost the same whether they come from one playlist read deep or five read to their free limit, because four extra allowances of a hundred exactly offset four extra nickels. Which one you want depends on the question, not on the bill.

Who works from a curated list?

The listing names course research, content auditing and video curation. A playlist is someone else’s editorial judgement already applied, which is exactly what makes it cheaper to analyse than a whole channel.

• Course and training research

  One playlist

  Sum duration across a curriculum to get its real running time before committing anyone to it

  $0.05 per course

• Content auditing

  Several lists

  Check a whole shelf of playlists for dead, private or livestream entries in one pass

  $0.25 per 500 videos

• Music and catalogue work

  One playlist

  Read an album or a label playlist with numeric views on every track

  $0.05 per 100 tracks

• Curation research

  Find playlists

  Search for how others have grouped a topic, at the same price as naming a list outright

  $0.05 per search

• Transcription pipelines

  One playlist, then pivot

  Collect video IDs from a list, then hand them to the Transcript Scraper

  $0.05 per 100 IDs

Reading a list end to end

The actor is apidojo/youtube-playlist-scraper. Since the allowance covers most real playlists entirely, the useful patterns are about what you do with the list rather than how you page through it.

from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")

run = client.actor("apidojo/youtube-playlist-scraper").call(run_input={
    "startUrls": ["https://www.youtube.com/playlist?list=PL…"],
    "maxItems": 100,     # 100 come free, so most playlists cost one query fee
})

videos = list(client.dataset(run.default_dataset_id).iterate_items())

# duration is a guaranteed number, so a curriculum's length is one sum.
watchable = [v for v in videos if not v["isLive"] and not v.get("isPrivate")]
total_min = sum(v["duration"] for v in watchable) / 60
print(f"{len(watchable)} videos, {total_min:.0f} minutes")

# description and keywords are the only conditional fields, so guard them.
tagged = [v for v in watchable if v.get("keywords")]
print(f"{len(tagged)} of {len(watchable)} carry tags")

for v in sorted(watchable, key=lambda v: v["views"], reverse=True)[:5]:
    print(f"{v['views']:>12,}  {v['duration']:>5}s  {v['title'][:50]}")

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });

// One run per playlist: the cost is identical and the provenance survives.
const lists = ['PLaaa…', 'PLbbb…'];

for (const id of lists) {
  const run = await client.actor('apidojo/youtube-playlist-scraper').call({
    startUrls: [`https://www.youtube.com/playlist?list=${id}`],
    maxItems: 100,
  });
  const { items } = await client.dataset(run.defaultDatasetId).listItems();

  const dead = items.filter((v) => v.isPrivate);
  const hours = (items.reduce((n, v) => n + v.duration, 0) / 3600).toFixed(1);
  console.log(id, items.length, 'videos', hours + 'h', dead.length, 'private');
}

curl -X POST \
  "https://api.apify.com/v2/acts/apidojo~youtube-playlist-scraper/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"startUrls":["https://www.youtube.com/playlist?list=PL…"],"maxItems":100}'

Export as JSON, CSV or Excel from the Storage tab. JSON keeps the thumbnails and keywords arrays intact, and CSV is fine if all you want is the durations and view counts.

YouTube Playlist Scraper FAQ

Where YouTube Playlist Scraper fits