TikTok Video Transcript Scraper

🎬 TikTok Video Transcript Scraper

Turn TikTok videos into searchable text. The TikTok Video Transcript Scraper returns the spoken words of a video — full transcript plus timestamped cues — alongside the whole post record: caption, hashtags, author, follower count, publish date, length, sound, and play, like, comment, share and save counts.

Give it a username and it pulls that account's recent videos on every run, which makes it a monitor rather than a one-off lookup. Paste video URLs instead when you already know exactly which ones you want.

✨ Features

• 🗣️ Transcript with timestamps — the full text plus cue-level start/end seconds, ready to search, quote or feed to an LLM.
• 👤 Track accounts, not just links — enter usernames and get their latest videos each run. Add a date filter and a scheduled run only ever returns what's new.
• 🌍 Any language — transcripts come back in whatever was spoken. English, Japanese, Russian, Spanish, Arabic and more, detected automatically.
• 🤖 AI transcription for the rest — most videos with speech already carry a transcript. Switch on AI transcription and the ones that don't get transcribed too, taking coverage to roughly 9 in 10.
• 📊 The whole post, not just the words — caption, hashtags, publish date, length, cover image, verified status, follower count, sound name, and five engagement counts.
• 🎯 Never pays for empty rows — photo posts and unavailable videos are skipped, and a video with no speech is charged as a video, not as a transcript.
• ⚡ No account needed — no login, no cookies, nothing to configure.

🛠️ How to Use the TikTok Video Transcript Scraper

1. Enter usernames – add one or more TikTok accounts (nasa, @bbc, or a profile URL).
2. Or paste video URLs – use these instead when you have specific videos in mind.
3. Set a date filter (optional) – 7 days or 2026-08-01 keeps a scheduled run to fresh posts only.
4. Run it – transcripts and post data arrive in the dataset, ready to export as JSON, CSV or Excel.

🔧 Input

Field	Type	Description
usernames	array	TikTok accounts to pull recent videos from. Username, @handle or profile URL.
videoUrls	array	Specific video URLs to transcribe. Use instead of, or alongside, usernames.
onlyPostsNewerThan	string	Skip anything older. Accepts 2026-08-01 or 7 days. Empty means no filter.
transcribeWithAI	boolean	Also transcribe videos that have no transcript of their own. Off by default.
language	string	Spoken language for AI transcription. Automatic by default.
maxResults	integer	Stop after this many videos. Default 50.
proxyConfiguration	object	Used only if TikTok rate-limits the run. Residential recommended.

📊 Sample Output Data

[
    {
        "id": "7665075736742530317",
        "url": "https://www.tiktok.com/@nasa/video/7665075736742530317",
        "handle": "nasa",
        "description": "Something big just landed on TikTok.",
        "hashtags": [],
        "createdAt": "2026-07-21T20:09:02.000Z",
        "durationSeconds": 25,
        "coverImage": "https://p16-common-sign.tiktokcdn-eu.com/tos-useast5-p-0068-tx/oAUhkWfCZC7g6A...",
        "authorId": "7664638705177150477",
        "authorName": "NASA",
        "authorVerified": true,
        "authorFollowers": 1200000,
        "authorBio": "Making the seemingly impossible, possible.✨",
        "playCount": 939200,
        "likeCount": 71000,
        "commentCount": 2038,
        "shareCount": 2717,
        "saveCount": 5430,
        "musicTitle": "original sound",
        "musicAuthor": "NASA",
        "musicIsOriginal": true,
        "detectedLanguage": null,
        "hasTranscript": true,
        "transcriptSource": "ai",
        "transcriptLanguage": "en",
        "transcript": "We're going back to the freaking moon. That's why.",
        "transcriptCues": [{ "start": 0, "end": 2.72, "text": "We're going back to the freaking moon. That's why." }],
        "skippedReason": null,
        "scrapedAt": "2026-08-22T09:15:26.398Z"
    }
]

📋 Data Fields

Field	Description
id, url, handle	Video identity and where it lives.
description, hashtags	The caption, and the hashtags pulled out of it.
createdAt, durationSeconds, coverImage	When it was posted, how long it runs, and its cover frame.
authorId, authorName, authorVerified, authorFollowers, authorBio	Who posted it and how big their audience is.
playCount, likeCount, commentCount, shareCount, saveCount	Engagement, all five counts.
musicTitle, musicAuthor, musicIsOriginal	The sound used, and whether it is the creator's own.
transcript	The spoken words as one block of text.
transcriptCues	The same text split into cues with start and end seconds.
transcriptLanguage, transcriptSource	Language spoken, and whether the text came from the platform or from AI transcription.
hasTranscript, skippedReason	Whether text was found, and in plain words why it wasn't.
detectedLanguage, scrapedAt	Speech language TikTok itself detected, and when this row was collected.

💵 Pricing

This Actor uses the pay-per-event pricing model — you only pay for what you get:

Event	Price	When it's charged
Actor start	$0.001	Once per run.
Video scraped	$0.003	Per video returned, with all its metadata.
Transcript returned	$0.005	Only when the video actually comes back with text.
Minute transcribed by AI	$0.020	Per started minute, only when AI transcription is used.

A video that comes back with a transcript costs $0.008; one with no speech costs $0.003. A typical run of 100 videos with transcripts on about two-thirds of them comes to roughly $0.63. Photo posts and unavailable videos are skipped and never charged. New Apify accounts include free monthly usage credits, so you can try it at no cost.

💡 Tips

• Start with a small maxResults to preview the data shape before scaling up.
• Pair usernames with onlyPostsNewerThan: 7 days on a weekly schedule to build a rolling archive of an account's talking points.
• Leave AI transcription off for a first pass. Most videos with speech already carry a transcript, and you can rerun the gaps later.
• Feed transcript into an LLM to summarise a competitor's messaging, extract hooks, or find every video where a product was mentioned.
• Use transcriptCues when you need to jump to the moment something was said rather than read the whole thing.
• Leave language on automatic unless a specific accent is being detected wrong.

⚠️ Good to Know

• Around 12 recent videos are available per account. This Actor is built to watch accounts over time, not to backfill an entire posting history in one run. For older videos, pass their URLs directly.
• Roughly 2 in 3 videos with speech already carry a transcript. Switching on AI transcription covers most of the remainder — together about 9 in 10.
• Videos with no spoken words — dance, music and wordless comedy — return no transcript, and skippedReason says so.
• TikTok limits how quickly one IP can read. For larger runs, leave the proxy enabled in the input.

❓ FAQ

Do I need a TikTok account or login? No. Nothing to log into and nothing to configure.

Can I get every video an account has ever posted? Not from a username alone — about 12 recent videos are listed per account. If you have the URLs of older videos, pass them in videoUrls and they will be transcribed.

Why did some videos come back without a transcript? Either there was no speech to transcribe, or the platform published no transcript for it and AI transcription was left off. The skippedReason field tells you which.

What languages are supported? Whatever was spoken. Transcripts have come back in English, Japanese, Russian, Spanish, Arabic, French, German and more, detected automatically.

Can I get comments? Not in this Actor — it focuses on the video and its transcript.

Is scraping TikTok legal? Collecting publicly available data is generally lawful in the EU and US. Use the output responsibly and respect copyright and personal data rules in your jurisdiction.

🆘 Support

Found a bug or need a field that isn't here? Open an issue on the Actor's Issues tab and it will be looked at.