TikTok Transcript Scraper & Extractor - AI Video to Text

TikTok transcript scraper and extractor: turn TikTok videos into text. This Actor transcribes the spoken audio of public TikTok videos with AI speech-to-text, so it works even when a video has no captions. Paste video URLs or just a creator's username, get clean transcripts with views, likes, comments and shares as JSON, CSV or Excel, plus SRT and WebVTT subtitle files, or pull them into your app through the API. No TikTok login needed.

Why this TikTok transcript tool is different

Many TikTok videos are voice-overs on top of a popular song. TikTok exposes the song as a separate "music" track, and transcribing that track returns the song lyrics (or nothing) instead of what the creator said. This Actor never uses the music track: it transcribes the audio of the video itself, so you get the creator's actual words.

What it does

• Transcribe TikTok videos to text, including videos without captions
• Transcribe a creator's newest videos from just their username or profile URL (pinned posts skipped)
• Return the full transcript, word count and audio duration for every video, with its post date, plays, likes, comments and shares
• Optionally return timestamped segments, the detected spoken language and ready-to-use SRT and WebVTT subtitles
• Accept full links (tiktok.com/@user/video/...) and short links (vm.tiktok.com/...)
• Process many videos in one run and skip duplicate URLs automatically
• Retry region-restricted videos through the proxy
• Stop exactly at the spending limit you set for a run

Who uses TikTok video to text?

• Content and social media teams turning TikToks into posts, captions and scripts
• Marketers and researchers analysing what creators and competitors say
• Developers and AI agents feeding transcripts into summarisers, classifiers, search indexes or RAG pipelines
• Accessibility work that needs a text version of short videos

How to transcribe TikTok videos

1. Add one or more public TikTok video URLs.
2. Optional: set the spoken language (for example en, es, pt, ko) for short or noisy clips.
3. Optional: turn on Include segment timestamps.
4. Click Start. Each video appears in the dataset as soon as it is done.

Transcribe all videos of a TikTok profile

Put usernames or profile URLs in Profiles instead of (or next to) video URLs:

{
  "profiles": ["nasa", "https://www.tiktok.com/@natgeo"],
  "maxVideosPerProfile": 20,
  "includeTimestamps": true
}

The newest videos of each profile are listed with the TikTok Profile Scraper by Clockworks, then transcribed. Real row from a test run:

{
  "url": "https://www.tiktok.com/@nasa/video/7695090484124929293",
  "profile": "nasa",
  "postedAt": "2026-10-10T17:21:28.000Z",
  "plays": 141400,
  "likes": 7801,
  "comments": 277,
  "shares": 289,
  "status": "succeeded",
  "durationSec": 122.4,
  "text": "This week, NASA's missions stretched from the edge ...",
  "srt": "1\n00:00:00,000 --> 00:00:04,320\nThis week, NASA's missions stretched from the edge of the Arctic ...",
  "charged": true
}

Profiles that are private, do not exist, or have no videos in the date range are returned as a free failed row with errorCode PROFILE_NOT_LISTED.

TikTok transcript generator and TikTok transcript API

Use it two ways. As a TikTok transcript generator: paste TikTok links in Apify Console, click Start and download the transcripts as TXT, CSV, JSON or Excel. As a TikTok transcript API: call the Actor from your code or AI agent and get JSON back (see the API example below). Both work in bulk, from single videos to a creator's whole profile.

How to get the transcript of a TikTok video

1. In TikTok, open the video, tap Share and Copy link (short vm.tiktok.com links work too).
2. Paste the link into TikTok video URLs, or put the creator's username in Profiles to get all their newest videos.
3. Click Start. The transcript, word count, views, likes, comments and shares appear in the dataset within seconds.

Unlike caption-only tools, this works on TikToks without captions or subtitles: the creator's spoken audio is transcribed with AI speech-to-text, and the background song track is ignored.

Output example

Real output from a test run:

{
  "url": "https://www.tiktok.com/@pokemonlife22/video/7059698374567611694",
  "platform": "tiktok",
  "videoId": "7059698374567611694",
  "status": "succeeded",
  "durationSec": 6,
  "text": "Japanese boxes, if anybody's interested, let me know.",
  "wordCount": 8,
  "postedAt": "2022-02-01T11:15:23+00:00",
  "plays": 389,
  "likes": 40,
  "comments": 2,
  "shares": 0,
  "model": "gpt-4o-mini-transcribe",
  "charged": true
}

Field	Description
status	succeeded, no_speech (music only), no_audio, failed or skipped
text, wordCount, durationSec	Transcript, its length, and audio length in seconds
language, segments, srt, vtt	Present when timestamps are enabled
author, caption	Video metadata when TikTok provides it
postedAt, plays, likes, comments, shares	Post date and engagement at processing time
profile, hashtags	Profile mode only: where the video was listed and its hashtags
errorCode, errorMessage	Why a video was not transcribed
charged	Whether the item was billed

Pricing

Pay per event, no subscription:

Event	Price
Video transcribed (first minute of audio)	$0.01
Each additional started minute	$0.01

1,000 TikToks of up to one minute cost $10; longer videos add $0.01 per extra started minute.

Not charged: invalid URLs, private or removed videos, videos TikTok blocks, and videos without an audio track. A video that contains only music is transcribed and returned as no_speech; it is billed like any transcription, because the audio was processed. Set a maximum charge per run and the Actor never exceeds it.

Profile mode: listing a profile's videos runs the TikTok Profile Scraper by Clockworks in your account, which Apify bills separately at about $0.003 per video listed ($0.0013 more per video when you use the date filter). A one-minute video from a profile therefore costs about $0.013 in total. The Actor never lists more videos than your run budget can transcribe.

Use it from the API and AI agents

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("<username>/tiktok-transcript").call(
    run_input={"urls": ["https://www.tiktok.com/@pokemonlife22/video/7059698374567611694"]}
)
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["status"], item["text"])

Works with schedules, webhooks, Make, Zapier and n8n. AI agents connected to the Apify MCP server can find and call it.

Limits and FAQ

• Public videos only; no login, no private accounts.
• Audio longer than Maximum minutes per video (default 10) is cut at that length.
• Does it read on-screen text? No, it transcribes what is spoken.
• Can I use the text with my own AI model? Yes; the output is plain JSON for your own pipeline.
• Is it legal? The Actor only processes publicly available videos and does not log in. You are responsible for how you use the data; check TikTok's terms and the laws that apply to you.