🎵 TikTok Trending Videos Scraper Plus

Extract trending TikTok videos by country, sort order and time period — or scrape any list of specific TikTok video URLs — with full metadata harvested from each video page. This TikTok scraper reads the video's embedded SIGI_STATE / rehydration JSON and maps every useful field (views, author details, music, hashtags, full caption, timestamps, video specs) instead of discarding it. It also computes engagement metrics, supports filtering, optional multi-provider AI enrichment, a real proxy, deduplication and canonical URLs.

Keywords: tiktok scraper, tiktok trending videos, tiktok data extractor, tiktok api, tiktok video scraper, tiktok hashtag data, tiktok analytics, tiktok creator metrics, tiktok music data, tiktok engagement rate.

What it does

Two modes:

1. Trending mode — pulls trending videos from TikTok's Creative Center for a country / sort / period.
2. Video-URL mode — set startUrls to scrape specific TikTok video links (reuses the same page parser). When startUrls is provided, country/sort/period/maxItems are ignored and detailed data is always harvested.

Output fields

Base (always): id, item_id, item_url, webVideoUrl, title, region, cover, country_code, duration, type, scrapedAt.

With Include Detailed Video Data (or in Video-URL mode):

• Views & engagement: views / playCount, likes, comments, shares, bookmarks, totalEngagement, engagementRate (%), likesPerView, commentsPerView, sharesPerView, bookmarksPerView.
• Author: authorId, authorUniqueId, authorNickname, authorVerified, authorFollowers, authorAvatar, authorBio, authorUrl.
• Music: musicId, musicTitle, musicAuthor, musicPlayUrl, musicDuration, musicOriginal.
• Content: desc (full caption), hashtags[] (from challenges + textExtra + regex fallback), createTime, createTimeISO, isAd.
• Video specs: videoWidth, videoHeight, videoDuration, videoRatio, videoDefinition, videoFormat, videoBitrate, videoUrl (expiring direct link).
• AI (optional): aiSentiment, aiTopics, aiKeywords, aiLanguage.

Filters (require Detailed Video Data)

minViews/maxViews, minLikes/maxLikes, minComments/maxComments, minDuration/maxDuration (seconds), verifiedOnly, publishedAfter/publishedBefore. Provide them as top-level quick-filter fields or inside the filters object.

AI enrichment (optional, off by default)

Set aiEnhancement: true, pick an aiModel, and supply aiApiKey (or set the matching provider env var). Provider is auto-detected from the model name: claude-* → Anthropic, gpt-*/o1/o3 → OpenAI, gemini-* → Google, grok-* → xAI, deepseek-* → DeepSeek, sonar* → Perplexity, mistral-* → Mistral. Env-var fallbacks: ANTHROPIC_API_KEY, OPENAI_API_KEY, GEMINI_API_KEY/GOOGLE_API_KEY, XAI_API_KEY, DEEPSEEK_API_KEY, PERPLEXITY_API_KEY, MISTRAL_API_KEY. If no key is available, AI enrichment is skipped with a warning (the run still succeeds).

Proxy

proxyConfiguration is wired into both the aiohttp HTTP client and the Playwright browser for the anti-bot-sensitive TikTok video-page fetches (proxy first, with a direct-connection fallback if the proxy IP is blocked). The TikTok Creative Center trending listing is fetched directly, because that ads endpoint does not route reliably through datacenter proxies. Apify Proxy (residential) is recommended for the video pages.

Example input

{
  "countryCode": "US",
  "sortBy": "vv",
  "period": "7",
  "maxItems": 20,
  "includeDetailedVideoData": true,
  "minViews": 100000,
  "verifiedOnly": false,
  "proxyConfiguration": { "useApifyProxy": true }
}

Video-URL mode:

{
  "startUrls": [
    "https://www.tiktok.com/@scout2015/video/6718335390845095173"
  ],
  "proxyConfiguration": { "useApifyProxy": true }
}

Billing

Charged per video row (row_result pay-per-event). Error rows (unparseable/blocked pages) are pushed unbilled for transparency.

Notes & limitations

• Engagement metrics and filters need the video page (Detailed Video Data / URL mode).
• videoUrl direct links are short-lived (TikTok CDN expiry).
• TikTok anti-bot may block some page fetches; use Apify Proxy and retry. Deferred (not built) in this version: comments/replies, author's other videos, related/music videos, media downloads, and hashtag/profile/search discovery.
• Only public data is scraped.