YouTube Scraper Actor (no external APIs) Scrape YouTube channels, playlists, videos, and search results using Playwright and YouTube’s own page data (DOM + window.ytInitial*). No YouTube Data API and no youtube-transcript-api. ## What’s new - Channel info only mode: set channelInfoOnly = true to scrape only the channel’s “About” data for the URLs you provide (channel or video). The actor will not scrape any videos or run searches in this mode. - Accepts: - Channel URLs (e.g., https://www.youtube.com/@theleaderstalk, https://www.youtube.com/channel/UCC...) - Video URLs (the actor resolves the channel from the video) - A single @handle in searchTerms (e.g., ["@theleaderstalk"]) Output in this mode (type: "channel_info"): - channelName - channelUrl (prefers handle URL when available) - channelUsername (handle without the @) - channelId - description - channelLocation - joinedDate - channelTotalVideos - channelTotalViews (numeric) and channelTotalViewsFormatted (with commas) - numberOfSubscribers (numeric) and subscriberCountText - socialLinks (array of { title, url }) - avatarUrl - isMonetized (best-effort) - sourceUrl (the input you provided) ## Other capabilities - Channel enrichment (when not in channel-only mode): - name, URL, ID, @handle, subscriber counts, total views, total videos, location, social links, avatar, description, joined date - Video details: - title, URL, thumbnail, view count, likes (best-effort), comments count (best-effort) - release date (ISO), duration (seconds + HH:MM:SS), hashtags - optional subtitles scraped via the UI, optional comments - Search with rich UI filters (applied via YouTube “Filters” menu): - sorting, date filter, type, duration - feature toggles: HD, Subtitles/CC, Creative Commons, 3D, Purchased, 4K, 360°, Location - Per-term limits for search (videos/shorts/streams) ## Input Key fields: - channelInfoOnly (boolean): Only scrape channel About info for provided URLs. If true: - Provide one or more channel or video URLs in startUrls, or - Provide a single @handle in searchTerms; the actor will treat it as a channel. - startUrls: URLs to process (channel, video, playlist, search results) - directUrls: extra URLs to merge into startUrls - searchTerms: terms to search (ignored when channelInfoOnly is true unless a single @handle is provided) - Subtitles options: downloadSubtitles, saveSubsToKVS, subtitlesLanguage, preferAutoGeneratedSubtitles, subtitlesFormat - Search filters (non-channelOnly): searchSortOrder, searchDateFilter, searchVideoType, searchDuration - Features (non-channelOnly): hd, subtitles_cc, creative_commons, three_d, purchased, four_k, three_sixty, location - Limits: maxVideosPerTerm, maxShortsPerTerm, maxStreamsPerTerm, maxResults, maxComments - Filters for post-processing: dateFilter, videoRegex - proxyConfiguration, verboseLog ### Examples Channel info only json { "channelInfoOnly": true, "startUrls": [ { "url": "https://www.youtube.com/@theleaderstalk" }, { "url": "https://www.youtube.com/watch?v=zslTD2Mq00E" } ], "proxyConfiguration": { "useApifyProxy": true } } Channel + videos (not channel-only) json { "startUrls": [{ "url": "https://www.youtube.com/@LofiGirl" }], "maxResults": 50, "proxyConfiguration": { "useApifyProxy": true } } Search with filters json { "searchTerms": ["traveler", "ai trends 2025"], "maxVideosPerTerm": 15, "maxShortsPerTerm": 8, "maxStreamsPerTerm": 2, "searchSortOrder": "relevance", "searchDateFilter": "this_year", "searchVideoType": "video", "searchDuration": "any", "hd": true, "subtitles_cc": true, "downloadSubtitles": true, "fetchFullVideoDetails": true, "maxResults": 30, "proxyConfiguration": { "useApifyProxy": true } } To search for the keyword json { "startUrls": [], "searchTerms": ["marketing"], "maxResults": 10, "maxVideosPerTerm": 10, "proxyConfiguration": { "useApifyProxy": false }, "verboseLog": false } ## Notes and limitations - YouTube does not expose subscriber usernames publicly; this actor cannot collect them. - Likes and comments counts can be hidden; values are best‑effort. - Layouts vary by region; enable Apify Proxy for more consistent results. ## Local development bash pip install -r requirements.txt python -m playwright install --with-deps chromium python main.py