Tiktok Scraper

TikTok Scraper

What is TikTok Scraper?

TikTok Scraper is an Apify Actor in this repo that collects public TikTok video metadata (caption, author, counts, music, cover, URL) from:

• Hashtags (hashtags → /tag/...)
• Profiles (profiles → /@...)
• Search keywords (searchQueries)
• Direct video URLs (videoUrls / startUrls)

It is built with Playwright in the same spirit as Amazon / Agency Vista actors here (Python + Chromium + Apify dataset).

Reference (feature & pricing model): the well-known Store Actor clockworks/tiktok-scraper. This CloudBots Actor is a separate implementation—not a fork of Clockworks’ code. TikTok changes often; for production scale many teams use the Store Actor or extend this one.

➡️ Provide seeds + limits; the Actor discovers video links on listing pages, opens each video, merges embedded JSON and XHR JSON, and pushes one row per video.

Why use this Actor?

• Own the code in your monorepo (customize fields, logging, proxy rules).
• Same deployment pattern as your other Actors (Dockerfile, INPUT_SCHEMA.json, dataset schema).
• Aligns input ideas with the Store scraper: hashtags, profiles, search, caps per seed, maxItems.

Features

• Seeds: hashtags, profiles, searchQueries, startUrls / urls, videoUrls.
• Discovery: scrolls listing pages, collects /video/ links + parses JSON from network and <script> embeds (__UNIVERSAL_DATA_FOR_REHYDRATION__, SIGI_STATE, etc.).
• Detail pass: per-video page with XHR capture + embedded JSON.
• Output: webVideoUrl, text, authorMeta, engagement counts, musicMeta, videoMeta, sourceType / sourceValue, error.
• Apify Proxy: default RESIDENTIAL when available (proxyConfiguration, disableProxy).
• Optional session: cookies (array) or storageState (Playwright JSON) so you can pass a logged-in browser export when TikTok shows login/captcha; treat values as secrets.

How to use on Apify

1. Push tiktok-scraper/ as an Actor (or connect Git subfolder).
2. Build from Dockerfile.
3. Set input (see below). Use RESIDENTIAL proxy and a long timeout for larger maxItems.
4. Start → Dataset.

Input (example)

{
  "hashtags": ["bananas"],
  "profiles": ["tiktok"],
  "searchQueries": ["ootd"],
  "resultsPerPage": 15,
  "maxItems": 25,
  "scrollRoundsPerSeed": 12,
  "proxyConfiguration": {
    "useApifyProxy": true,
    "apifyProxyGroups": ["RESIDENTIAL"]
  }
}

Local

pip install -r requirements.txt
playwright install chromium
python main.py

Uses INPUT.json when Apify input is empty (same pattern as other CloudBots actors).

Output

One dataset item per video, including:

• webVideoUrl, id, text, createTimeISO
• authorMeta: name, nickName, profileUrl, fans, etc.
• diggCount, playCount, commentCount, shareCount, collectCount
• musicMeta, videoMeta
• sourceType / sourceValue (which seed produced the link)
• error if that video failed after retries

Limitations

• Apify Proxy on SDK 3.x must use Actor.create_proxy_configuration(actor_proxy_input=...) (keyword-only). If you see takes 1 positional argument but 2 were given in logs, upgrade this Actor’s main.py—proxy was disabled and TikTok will often block.
• TikTok anti-bot is aggressive; expect blocks without working residential proxy and sober rate limits. Optional cookies / storageState help when you supply a valid session, but cookies expire and IP/proxy must still match what TikTok expects or the session may be rejected.
• Not feature parity with clockworks/tiktok-scraper (no video file downloads to KV in v0.1, no followers graph, etc.).
• DOM/API shapes change; empty datasets may need selector/JSON path updates.
• Respect TikTok ToS, GDPR, and Apify guidelines.

FAQ

How does this relate to clockworks/tiktok-scraper?

Same problem space and similar input concepts; different codebase. Use the Store Actor if you want Apify-maintained scale; use this if you want in-repo control.

Do I need an official TikTok API key?

No; this reads public web data like typical scrapers.

SEO Keywords

tiktok scraper apify
tiktok video metadata
tiktok hashtag scraper

Actor permissions

Designed for limited scope: default dataset + input only.

License

MIT (this Actor code).

Get started

Push to Apify, enable RESIDENTIAL proxy, start with small maxItems, then scale.