TikTok Real-Time Data

Collect live TikTok videos, creators, hashtags, sounds, suggestions, and comments as clean, structured JSON — streamed to your Apify dataset in real time.

Built for marketers, agencies, researchers, data teams, and AI pipelines that need reliable TikTok intelligence without building and maintaining a collection stack from scratch.

---

Why this Actor

	TikTok Real-Time Data	Typical DIY scrapers
Setup	Checkbox UI — click Start with prefilled keywords	Weeks of brittle scripts
Output	Structured JSON, one row per item	Messy HTML / mixed payloads
Scale	Long runs stream to the dataset (low RAM risk)	Often buffers everything in memory
Delivery	Dataset + optional real-time webhook	Manual export only
AI-ready	LLM / MCP friendly records	Ad-hoc parsing required
Coverage	Videos, creators, trends, suggestions, comments, URL paste	Usually one surface only

---

What you get

Every record includes featureType and scrapedAt so you can filter, join, and pipe into any workflow.

Videos (featureType: "video")

Field	Description
videoSource	How the video was found: keyword, url, related, user, hashtag, scrape
videoId, url, description, author, authorId	Identity
playCount, diggCount, commentCount, shareCount, collectCount	Engagement
createTime, durationSec, musicTitle, musicAuthor, hashtags[]	Media context
coverUrl, downloadUrl	Assets when available
searchQuery, position	Traceability
engagementFetched	Full engagement details were applied
accountFollowerCount, accountVerified, accountSignature, …	Optional creator fields on the same row

Creators (featureType: "account")

Field	Description
uniqueId, userId, nickname, url	Identity
signature, bioLink, avatarUrl, verified	Bio & trust
followerCount, followingCount, heartCount, videoCount	Reach
friendCount, privateAccount, isCommerce, language, createTime	Extra signals
discoveryQuery	Keyword that discovered this creator (keyword mode)

Hashtags & sounds

featureType	Key fields
hashtag	name, views, videoCount, description, coverUrl, url
sound	musicId, title, author, videoCount, coverUrl, url

Suggestions (featureType: "suggestion")

Field	Description
content	Suggested phrase / demand language
query	Seed keyword
isHashtag, position	Classification & rank

Comments (featureType: "comment")

Field	Description
commentId, text, author, authorId	Identity
diggCount, replyCount	Engagement
videoId, url, createTime, position	Parent video & order

Set includeRaw: true only if you need extended source fragments (larger output, advanced users).

---

Features

Keyword videos — on by default

Search by topic. Prefill: cooking, gym tips — click Start immediately. Each section has its own engagement, author details, and filters.

Videos by URL

Open specific TikTok video links or IDs, with optional enrichment toggles for this path.

Related videos

Seed video URLs → related recommendations, with path-level enrichment toggles.

Creator video grids

Recent posts from usernames (for full profiles, use Search or discover creators), with enrichment + min plays / date filters.

Search or discover creators

Two modes:

• By username / URL — deep profiles for known handles
• Search / discover by keyword — find creators matching a topic, then load full stats

Optional: also fetch each creator’s recent videos.

Hashtags & Sounds

• Discover — trending hashtags and sounds for your market
• Specific hashtags — meta for named tags, with optional videos under each tag

Search Suggestions

Autocomplete / related phrases TikTok associates with your keywords — ideal for demand research and content planning.

Comments

Three ways to pick videos: search by keyword, paste URLs, or reuse run-collected videos. Set videos per keyword, comments per video, and optional min comment likes.

Scrape by URL

Paste any TikTok video, profile, or hashtag URL — the Actor detects the type and returns the right record shape.

Webhooks

Every record is always saved to the run dataset. Optionally POST each record to your webhook URL (CRM, Slack, Zapier, Make, custom API) as it is collected.

---

Use cases

• Content & trend research — keyword videos with plays, likes, and music signals
• Creator discovery & outreach — find profiles by topic; export followers, bio, and bio link
• Brand monitoring — track hashtags, sounds, and mentions over time
• Competitive intelligence — related videos and creator grids around competitors
• Influencer shortlists — filter by engagement and account size
• Comment listening — sample audience language under high-performing videos
• AI & LLM pipelines — structured JSON for RAG, scoring, summaries, and draft content
• CRM / warehouse feeds — stream via dataset API or real-time webhook

---

LLM & MCP integration

Output is JSON Lines–friendly structured data — ideal for ChatGPT, Claude, Gemini, LangChain, LlamaIndex, and custom agents.

Recommended workflow

1. Run the Actor with the features you need (prefilled Videos works out of the box).
2. Fetch dataset items via Apify API or export JSON/CSV.
3. Pass records to your LLM with a system prompt, or index into a vector store.

Example: video record for an LLM prompt

{
  "featureType": "video",
  "videoSource": "keyword",
  "videoId": "7635410154141338893",
  "url": "https://www.tiktok.com/@creator/video/7635410154141338893",
  "description": "Easy weeknight dinner tips #cooking",
  "author": "creator",
  "playCount": 11900000,
  "diggCount": 850000,
  "commentCount": 1384,
  "searchQuery": "cooking",
  "scrapedAt": "2026-07-16T12:00:00.000Z"
}

Apify MCP (Model Context Protocol)

Use the Apify MCP server so AI assistants can:

• Run this Actor with natural-language instructions
• Read dataset results directly in the chat
• Chain with other Actors (e.g. collect → score → CRM)

Typical MCP tool flow:

User: "Find 30 cooking videos with high play counts and summarize hooks"
→ MCP runs Actor with enableVideoKeyword, videoKeywords=["cooking"], videoFetchEngagement=true
→ MCP reads dataset items
→ LLM summarizes descriptions, music, and engagement

API quick start

curl -X POST "https://api.apify.com/v2/acts/YOUR_ACTOR_ID/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "enableVideoKeyword": true,
    "videoKeywords": ["cooking", "gym tips"],
    "videoMaxResults": 20,
    "videoKeywordFetchEngagement": true
  }'

Dataset items: GET https://api.apify.com/v2/datasets/{datasetId}/items?format=json

---

Input reference

Enable only what you need. Features are independent. Defaults are prefilled so Start works immediately.

Input	Type	Default	Description
Keyword videos
enableVideoKeyword	boolean	true	Search videos by topic
videoKeywords	string[]	["cooking","gym tips"]	Search keywords
videoMaxResults	integer	20	Max videos per keyword
videoKeywordFetchEngagement	boolean	true	Full stats for keyword videos
videoKeywordFetchAccountDetails	boolean	false	Author fields on keyword rows
videoKeywordMinPlays / videoKeywordPublishedWithin			Filters for this path
Videos by URL
enableVideoUrl	boolean	false	Open specific video links
videoUrls / videoIds	string[]	—	Video URLs or IDs
videoUrlFetchEngagement / videoUrlFetchAccountDetails	boolean		Enrichment for URL path
Related videos
enableVideoRelated	boolean	false	Related / recommended clips
relatedVideoUrls	string[]	—	Seed video URLs
relatedMaxResults	integer	16	Related per seed
videoRelatedFetchEngagement / videoRelatedFetchAccountDetails	boolean		Enrichment for related path
Creator video grids
enableVideoUser	boolean	false	Recent posts from usernames
videoUsernames	string[]	—	Usernames without @
videoUserMaxResults	integer	20	Videos per creator
videoUserFetchEngagement / filters			Enrichment + min plays / date
Search or discover creators
enableAccounts	boolean	false	Look up or discover creator profiles
accountMode	enum	url	url or keyword
accountUsernames / accountUrls	string[]	—	Specific profiles
accountKeywords	string[]	—	Search / discover keywords
accountsPerKeyword	integer	15	Max creators per keyword
accountFetchRecentVideos	boolean	false	Also fetch recent posts
accountRecentVideosMax	integer	12	Videos per creator
Hashtags & Sounds
enableTrends	boolean	false	Trends section
trendsMode	enum	discover	discover, hashtag, all
hashtags	string[]	—	Specific tags (without #)
hashtagFetchVideos	boolean	true	Videos under each hashtag
hashtagMaxVideos	integer	20	Videos per hashtag
discoverMaxHashtags / discoverMaxSounds	integer	40 / 20	Discover caps
discoverHydrateDetails	boolean	true	Full hashtag stats on discover
Suggestions
enableSuggestions	boolean	false	Search suggestions
suggestionKeywords	string[]	—	Defaults to video keywords if empty
suggestionsMaxResults	integer	15	Per keyword
Comments
enableComments	boolean	false	Comment rows
commentKeywords	string[]	—	Find videos by keyword, then load comments
commentSearchMaxVideos	integer	5	Videos per comment keyword
commentsMaxResults	integer	30	Comments per video
commentMinDiggs	integer	0	Min likes on a comment (0 = off)
commentVideoUrls / commentVideoIds	string[]	—	Explicit video targets
commentsFromSearchResults	boolean	false	Also use videos from this run
commentsFromSearchMaxVideos	integer	10	Cap for run-collected videos
Scrape by URL
enableScrapeByUrl	boolean	false	Auto-detect URL type
scrapeUrls	string[]	—	Video / profile / tag URLs
Output & delivery
webhookUrl	string	—	Optional real-time POST per record
webhookFormat	enum	json	json or slack
includeRaw	boolean	false	Extended payloads

Full schema: .actor/input_schema.json or the Input tab on Apify Console.

---

Output reference

Each dataset row is one record. Filter by featureType:

featureType	Description
video	Unified video row (videoSource explains origin)
account	Creator profile
hashtag	Hashtag / trend tag
sound	Trending or discover sound
suggestion	Search suggestion phrase
comment	Single comment

Dataset views in Apify Console: Overview, Videos, Creators, Hashtags, Sounds, Comments.

Export: JSON, CSV, Excel, RSS, or API.

Streaming & memory

Records are pushed to the dataset as they are collected. The Actor does not hold the full run result set in RAM, so long runs stay stable. Control volume with the limits inside each feature section (videos per keyword, comments per video, etc.).

Webhook delivery

1. Set webhookUrl to your endpoint (HTTPS recommended).
2. Choose webhookFormat: JSON (full record) or Slack (short message).
3. Each successful dataset write is also POSTed to your URL (best-effort — webhook failures never drop dataset data).

---

Quick start examples

Keyword videos (default-style)

{
  "enableVideoKeyword": true,
  "videoKeywords": ["cooking", "gym tips"],
  "videoMaxResults": 20,
  "videoKeywordFetchEngagement": true
}

Search or discover creators by keyword

{
  "enableVideoKeyword": false,
  "enableAccounts": true,
  "accountMode": "keyword",
  "accountKeywords": ["gordon ramsay"],
  "accountsPerKeyword": 10,
  "accountFetchRecentVideos": true,
  "accountRecentVideosMax": 5
}

Trending hashtags & sounds

{
  "enableVideoKeyword": false,
  "enableTrends": true,
  "trendsMode": "discover",
  "discoverMaxHashtags": 30,
  "discoverMaxSounds": 15
}

Keyword videos + comments from this run + webhook

{
  "enableVideoKeyword": true,
  "videoKeywords": ["skincare"],
  "videoMaxResults": 10,
  "enableComments": true,
  "commentsFromSearchResults": true,
  "commentsFromSearchMaxVideos": 5,
  "commentsMaxResults": 20,
  "webhookUrl": "https://hooks.example.com/tiktok",
  "webhookFormat": "json"
}

---

Performance

• Connection routing is handled automatically by the Actor.
• Default memory: 1024 MB.
• Results stream to the dataset continuously; pair with webhooks for live pipelines.

---

Limitations & compliance

• Some surfaces (for example public comments) depend on what TikTok exposes without a user login; yield can vary by region and content type.
• Unofficial tool — not affiliated with TikTok. Use responsibly and comply with applicable laws and TikTok’s Terms of Service.

---

Contact & custom work

Need something beyond this Actor? I build custom scrapers, data pipelines, and full-stack web applications of any kind for startups and enterprises.

• Email: dubem115@gmail.com
• GitHub: github.com/DrunkCodes

Reach out for:

• Custom TikTok, Instagram, LinkedIn, Google Maps, or any site collection
• Lead-gen and enrichment systems
• Dashboards, internal tools, and production web apps
• AI agents and MCP-powered workflows wired to your data

---

License

ISC

---

TikTok Real-Time Data · by DrunkCodes