YouTube Scraper Intelligence
DeprecatedPricing
from $3.50 / 1,000 video scrapeds
Scrape public YouTube videos, Shorts, streams, channels, playlists, search results, subtitles, comments, and AI-ready transcript records.
Pricing
from $3.50 / 1,000 video scrapeds
Rating
0.0
(0)
Developer
Crowd Pull
Maintained by CommunityActor stats
2
Bookmarked
5
Total users
3
Monthly active users
4 months ago
Last modified
Categories
Share
Video, Shorts, channel, handle, playlist, or YouTube search-result URLs.
YouTube search terms. Use maxResults/maxResultsShorts/maxResultStreams to control output buckets.
YouTube handles or channel IDs. Examples: @bittensor, kurzgesagt, UC...
Scrape YouTube Trending for the selected gl/hl locale.
YouTube hl language parameter such as en, de, es, or ja.
Total output cap across all sources. If maxResults is omitted, this also sets the per-source regular-video target.
Maximum regular videos to emit per source.
Maximum Shorts videos to emit per source. Use 0 to exclude Shorts.
Maximum livestream or stream replay videos to emit per source. Use 0 to exclude streams.
Sorting order. Also accepts the sortingOrder alias with relevance, rating, date, or views.
Relative or custom upload date filter applied after extraction, and mapped into search URLs where supported.
Upload date alias: all, hour, today, week, month, or year.
Search-compatible type filter. Movie matching is best-effort from public metadata.
Single feature or comma-separated features: any, 4k, hd, live, movie, cc, 3d, hdr, location, purchased, subtitles, vr180, vr360.
Search-compatible length filter.
Duration alias. Maps short to under4, medium to between420, and long to plus20.
Only include videos with public subtitle tracks when true.
Only include videos marked with a Creative Commons license when true.
Compatibility-only purchased/premium filter. Public unauthenticated YouTube often returns no real video rows for this filter, and purchase availability is not reliably exposed.
YYYY-MM-DD or relative values such as 1 day, 7 days.
Post-sorts extracted channel/search candidates by newest, popular, or oldest.
Fetch channel about pages once per channel to fill description, links, subscriber count, total videos, and total views where public.
Extract public manual captions first and auto-generated captions second.
Compatibility alias. Enable this when downstream automations need stable KV URLs.
Language codes or regex patterns, tried in order. Example: en.*,es,fr.
[ "en.*"]Compatibility alias. Use any, en, de, es, fr, it, ja, ko, nl, pt, or ru.
Prefer auto-generated captions before manual captions.
Choose one subtitle payload format or keep all AI-ready formats.
Fetch comments with yt-dlp. Large values can significantly increase runtime.
Comment sorting mode passed to yt-dlp when comments are enabled.
Maximum comments to extract per video when includeComments is enabled.
Use captions only, or fall back to Whisper when no suitable caption track exists.
Total audio minutes allowed for Whisper fallback in one actor run.
Maximum individual video duration eligible for Whisper fallback.
Skip previously emitted videos per source until they age beyond the refresh window.
Recently seen videos older than this many days are refreshed so mutable metrics stay current.
JavaScript function. Receives each output row as item/input and should return JSON-serializable data.