Social Video Transcript: TikTok, Reels, YouTube, X & Facebook

Get the transcript of Instagram Reels, TikTok, Facebook, X (Twitter) and YouTube videos from their URLs: full text, timestamped segments, detected language and duration. Uses the platform captions when they exist, Whisper large-v3 AI otherwise. No login, no API key. Pay per video.

PNDA · Social media suite • Instagram, TikTok, Facebook, X, Reddit, Pinterest, Threads

Instagram Scraper Profiles, posts, reels, comments TikTok Scraper Videos by keyword or profile, comments Facebook Scraper Page & group posts, comments, reels X (Twitter) Scraper Tweets and profiles, no API key Reddit Scraper Posts by keyword or subreddit, users Pinterest Scraper Pins and profiles Threads Scraper Meta Threads posts and profiles Video Transcript ➤ You are here

What you get

Network	Example URL	How it is transcribed
Instagram Reels	https://www.instagram.com/reel/...	Whisper large-v3 AI
TikTok	https://www.tiktok.com/@user/video/...	TikTok captions when present, AI otherwise
YouTube (videos and Shorts)	https://www.youtube.com/watch?v=...	YouTube captions when present, AI otherwise
X (Twitter)	https://x.com/user/status/...	Whisper large-v3 AI
Facebook videos and reels	https://www.facebook.com/reel/...	Whisper large-v3 AI

One row per video:

{
  "url": "https://www.instagram.com/reel/DfBUU0nDswI",
  "network": "instagram",
  "status": "success",
  "transcriptSource": "ai-whisper",
  "language": "en",
  "durationSeconds": 40.92,
  "transcribedSeconds": 41.08,
  "truncated": false,
  "text": "I think by far the scariest thing in the universe is a black hole. You want to avoid them...",
  "wordCount": 123,
  "title": "Avoiding black holes is pretty easy on Earth...",
  "author": "natgeoscience",
  "publishedAt": "2026-10-28T13:00:02.000Z"
}

With timestamped segments on, each row also carries segments (start, end, text).

Videos that are private, deleted or without audio come back with status: "failed" and a reason, and are not charged.

How to use it

1. Paste the video URLs in Video URLs (mix networks freely).
2. Optional: force a language (otherwise detected), turn off timestamped segments, cap the max duration per video.
3. Run, then export the dataset as JSON, CSV or Excel.

Input example:

{
  "urls": [
    "https://www.instagram.com/reel/DfBUU0nDswI",
    "https://x.com/nasa/status/2114463017043624059"
  ],
  "includeSegments": true,
  "maxDurationMinutes": 30
}

Pricing (pay per event)

Event	Price
Video transcribed by AI, first minute of audio included	$0.0126 ($12.60 / 1,000 videos)
Each extra minute of audio beyond the first	$0.0048
Video transcript read from the platform captions (YouTube, TikTok)	$0.005 ($5 / 1,000 videos)
Actor start (once per run)	$0.01

Examples: a 30 s Reel = $0.0126; a 45 s TikTok with captions = $0.005; a 3 min X video = $0.0126 + 2 × $0.0048 = $0.0222. The run never spends more than your max charge per run: videos that the remaining budget cannot pay for are skipped, and long audio is cut to what the budget covers (truncated: true).

Use cases

• Content research: read what top creators say in their Reels and TikToks without watching them.
• Repurposing: turn videos into blog posts, captions, threads or newsletters.
• Brand and ad monitoring: search spoken mentions of a brand or product across networks.
• Datasets for AI: build text corpora from short-form video.

FAQ

Which language? Whisper detects about 100 languages automatically; set language to force one.

How accurate is it? Whisper large-v3 is OpenAI's most accurate open speech model. Platform captions are used as-is when they exist (cheaper and instant).

Is it legal? The actor only reads public videos, like a visitor without an account. You are responsible for how you use the transcripts (copyright, personal data under GDPR / CCPA).

Free plan? Apify Free accounts can test with a few short videos per run.