YouTube Channel Videos Scraper: a channel’s uploads

Name the channel, or go looking for one

Same eight fields as the Channel Scraper, different prices on all of them. A channel query is the expensive one and the one that actually returns videos; a keyword search returns channels and costs five times what the same search costs on the sibling actor.

• startUrls

  $0.015per channel

  ~30 videos in

  Channel URLs. The main door. This is what triggers the $0.015 channel query that returns uploads.

  https://www.youtube.com/@ThreeDaysGraceVEVO

• youtubeHandles

  $0.015per channel

  ~48 if Shorts

  The @handle form, when you would rather store handles than URLs. Billed the same as a channel URL.

  ["@MrBeast"]

• Do your searching on the other actor

  keywords

  $0.005per search

  ~20 channels in

  A term to search channels with. Cheaper than a channel query but five times what the Channel Scraper charges for it.

  ["acoustic covers"]

• gl / hl / sort

  us / en / rthe defaults

  Search only

  Country, language, and ordering. The schema notes sort works only with search keywords.

  sort: "r"

maxItems and the "maximum charged results" field work together, and the listing warns that leaving both empty produces infinite results, which against a channel with ten thousand uploads is a five-dollar run you did not intend.

One catalogue, a set of channels, or the files themselves

Three runs. The third is not a different input. It is the same run read differently, because the streaming block is already in every record you have paid for.

{
  "startUrls": ["https://www.youtube.com/@ThreeDaysGraceVEVO"],
  "maxItems": 100
}

{
  "id": "lL2ZwXj1tXM",
  "title": "Three Days Grace - Never Too Late",
  "status": "OK",
  "url": "https://www.youtube.com/watch?v=lL2ZwXj1tXM",
  "duration": 211,
  "views": 276328963,
  "likes": 1796037,
  "isLive": false,
  "isPrivate": false
}

{
  "youtubeHandles": ["@MrBeast", "@ThreeDaysGraceVEVO"],
  "maxItems": 500
}

{
  "channel": {
    "id": "UCK-VRwguZP3cokOAxI0F1HA",
    "name": "ThreeDaysGraceVEVO",
    "url": "https://www.youtube.com/channel/UCK-VRwguZP3cokOAxI0F1HA"
  },
  "keywords": ["three days grace", "rock", "official video"]
}

{
  "startUrls": ["https://www.youtube.com/@ThreeDaysGraceVEVO"],
  "maxItems": 30
}

{
  "streamingData": {
    "adaptiveFormats": [
      {
        "itag": 140,
        "mimeType": "audio/mp4; codecs=\"mp4a.40.2\"",
        "bitrate": 130517,
        "audioQuality": "AUDIO_QUALITY_MEDIUM",
        "audioSampleRate": "44100",
        "audioChannels": 2,
        "contentLength": "3418563"
      }
    ]
  }
}

A video, and the file behind it

Four blocks, and one of them is far larger than the rest. streamingData routinely dwarfs everything else in a record, which matters when you are deciding what to keep.

• 8

  The video

  Top level

  • id, title, url
  • status, "OK" on a healthy record
  • description, full, untruncated
  • duration, seconds, integer
  • isLive, isPrivate

• 2

  Engagement

  Real numbers

  • views, 276328963
  • likes, 1796037
  • No parsing required

• 3

  The uploader

  channel

  • id, UC format
  • name
  • url

• 2

  Media

  The bulky part

  • thumbnails, sizes from 120px up
  • streamingData.formats
  • streamingData.adaptiveFormats

adaptiveFormats is where a player looks and where you should too: it separates video from audio, so a job that only needs the soundtrack can take a three-megabyte audio stream instead of the full file.

Before you run it against a big channel

Three from the listing’s own pricing and troubleshooting, and two that follow from what the output actually contains.

• 01

  Both ceilings default to unlimited

  The troubleshooting section names maxItems alongside "maximum charged results" and warns that leaving both empty results in infinite results. Against a ten-thousand-upload channel that is a five-dollar run.

• 02

  The free page is not a fixed number

  It is roughly 48 Shorts or roughly 30 videos. Budget on thirty and treat anything more as a discount rather than planning around the higher figure.

• 03

  Search here costs five times what it costs next door

  A keyword search is $0.005 on this actor and $0.001 on the Channel Scraper, both including about twenty channels. Find channels there, pull uploads here.

• 04

  Stream URLs are short-lived

  The googlevideo links carry expiry timestamps and signatures. Treat them as fetch-now, not as archival references.

• 05

  Records are large

  streamingData carries an entry per rendition with codecs, byte ranges and URLs. If you only need engagement, strip it with customMapFunction, which the listing permits for transformation but bans for filtering.

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

Where the channel fee stops mattering

Six scenarios from the listing, all reproducing from the three charge events. The channel query is the largest single fee in the YouTube family, and it stops being the dominant part of a bill at about a hundred videos.

The fourth and fifth rows are the same price for the same videos, which is the clearest signal in this pricing: below a few hundred videos per channel, spreading a run across more channels is free. Above it, the item rate takes over and the allowances stop helping.

Who needs the uploads themselves?

The listing points at influencer analysis and lead generation, but the streaming block widens that considerably, anyone who needs the media and not just the metrics is here rather than on the channel or comments actors.

• Content analysis

  One catalogue, deep

  Read a channel’s entire output with numeric views and likes that need no parsing before comparison

  $0.25 per 500 videos

• Competitive benchmarking

  A set of channels

  Compare cadence and performance across five channels for the price of pulling one

  $0.25 across 5 channels

• Media and archival work

  The files

  Take the streaming formats to fetch renditions directly, choosing bitrate and codec before downloading

  $0.015 per channel page

• Transcription pipelines

  The files, audio only

  Pull the audio-only adaptiveFormat rather than the full video, then hand it to a speech model

  $0.05 per 100 videos

• SEO and tag research

  One catalogue

  Mine the keywords array across a channel’s uploads to see how it actually tags itself

  $0.50 per 1,000 videos

Taking what you need and dropping the rest

The actor is apidojo/youtube-channel-scraper. Because streamingData dominates the payload, the first decision is whether you want it, and if you do, use it quickly.

from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")

run = client.actor("apidojo/youtube-channel-scraper").call(run_input={
    "startUrls": ["https://www.youtube.com/@ThreeDaysGraceVEVO"],
    "maxItems": 100,      # never leave this empty on a large channel
})

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

# views and likes are integers here: no parsing, unlike the comments actor.
live_excluded = [v for v in videos if not v["isLive"] and not v["isPrivate"]]
best = sorted(live_excluded, key=lambda v: v["likes"] / max(v["views"], 1), reverse=True)

for v in best[:10]:
    rate = v["likes"] / v["views"] * 100
    print(f"{rate:.2f}%  {v['duration']:>5}s  {v['title'][:50]}")

# Audio-only renditions are a fraction of the size of the full file.
def audio_stream(v):
    fmts = v.get("streamingData", {}).get("adaptiveFormats", [])
    return next((f for f in fmts if f.get("mimeType", "").startswith("audio/")), None)

import { ApifyClient } from 'apify-client';

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

const run = await client.actor('apidojo/youtube-channel-scraper').call({
  youtubeHandles: ['@MrBeast', '@ThreeDaysGraceVEVO'],
  maxItems: 500,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();

// channel.id keeps a multi-channel run sorted.
for (const [id, vids] of Object.entries(Object.groupBy(items, (v) => v.channel.id))) {
  const views = vids.reduce((n, v) => n + v.views, 0);
  const mins = Math.round(vids.reduce((n, v) => n + v.duration, 0) / 60);
  console.log(vids[0].channel.name, vids.length, 'videos', views.toLocaleString(), 'views', mins, 'min');
}

// Stream URLs expire, fetch now or re-run later, do not archive the link.
const smallest = items[0]?.streamingData?.adaptiveFormats
  ?.filter((f) => f.contentLength)
 .sort((a, b) => Number(a.contentLength) - Number(b.contentLength))[0];
console.log('smallest rendition:', smallest?.mimeType, smallest?.contentLength);

curl -X POST \
  "https://api.apify.com/v2/acts/apidojo~youtube-channel-scraper/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"startUrls":["https://www.youtube.com/@ThreeDaysGraceVEVO"],"maxItems":30}'

Export as JSON, CSV will flatten the streaming and thumbnail arrays into something unusable. If the streams are not needed, strip them in customMapFunction and the payload shrinks by an order of magnitude.

YouTube Channel Videos Scraper FAQ

Where YouTube Channel Videos Scraper fits