TikTok Transcript Scraper

Paste TikTok video links, get what is said in each video as clean text — with timestamps and language, straight from TikTok's own captions.

If you study what works on TikTok, you know the problem. The value of a video is in what the creator says — the hook, the script, the product claim — and none of it is in the caption. So you watch, pause, rewind and type. For ten videos that's an afternoon; for a competitor's last 200 videos it never happens. And pasting a link into a generic transcription tool means waiting for audio uploads and paying per minute.

We built this for our own content research first, and now we're sharing it. Most talking videos on TikTok already carry an automatic caption track that TikTok generated itself. This Actor reads that track directly — no audio download, no AI guessing — so a transcript costs a tenth of a cent and arrives in about a second. You get the full text, every line with its start and end time, the language, and the video's caption, creator and stats alongside.

Who it's for

• Content marketers and creators studying hooks and scripts — "what do the top 50 videos in my niche actually say in the first 3 seconds?"
• Agencies repurposing TikToks into blogs, newsletters and shorts — "give me the text of our client's last 30 videos."
• Brand safety and compliance teams — "did the creator say the required disclosure out loud?"
• Researchers and AI builders — transcripts as clean JSON for summaries, topic tagging, sentiment or RAG.

What you get per video

• transcript — the full spoken text; wordCount.
• segments — every caption line with start / end in seconds (turn off with Include timed segments).
• language (TikTok's tag, e.g. eng-US), languageCode (en), isAutoGenerated, transcriptSource (tiktok-captions or speech-to-text), availableLanguages.
• The video: URL, creator, caption, hashtags, post date, length, views, likes, comments, shares, sound.

Options

• Video URLs — full links, share links (vm.tiktok.com, vt.tiktok.com, tiktok.com/t/…) or bare video IDs, one per line. Up to 5,000 per run.
• Preferred language — optional (en, es, de…). If TikTok has captions in that language for a video you get them, otherwise the original language; the row tells you which.
• Include timed segments — on by default.
• Transcribe videos without captions — speech-to-text (Whisper) for videos TikTok didn't caption, when available for this Actor; with a length limit you set. Off, or not available: those videos come back as no-captions and cost nothing. If the speech-to-text service is busy (rate limit), the video comes back as stt-rate-limited, free — just run it again later.

How to set it up

1. Paste your video links into Video URLs.
2. Click Start. A hundred videos take about a minute.
3. Open the Transcripts view, export as CSV / Excel / JSON, or pull the rows into your tool or LLM pipeline through the API.

And that's it.

Pricing

• $0.002 per transcript from TikTok's captions — $2 per 1,000 videos.
• $0.005 per speech-to-text transcript (per started 5 minutes of audio), only for videos without captions and only when a transcript is delivered.
• Plus Apify's small start fee per run. Videos without captions (when speech-to-text is off or unavailable), videos without speech, deleted or private videos, invalid links and duplicates are free.

Tip: TikTok captions most videos where someone talks. Music-only videos, dances and very old videos usually have no captions — keep speech-to-text off if you only want the cheap ones.

Notes

• Transcripts are TikTok's own automatic captions, so they read the way TikTok's captions read on screen: very good for clear speech, weaker for heavy slang, music over speech or several people talking at once.
• Public videos only. No login, nothing liked or followed.
• Respect creators' rights when you reuse their words, and TikTok's terms.

Found a bug or need another field? Open an issue on the Issues tab. Need transcripts for a whole account or hashtag? Ask there too.