YouTube Transcript & Metadata Extractor (LLM-ready) 🔥🎬

Extract video transcripts (subtitles/captions) and detailed metadata from YouTube URLs.

Reads input 📥 and writes structured transcript data 📦.

Common Use Cases 💡

• Feed full transcripts into LLMs (summaries, embeddings, RAG)
• Analyze YouTube content at scale
• Build SEO or content research pipelines
• Power n8n / Make / Zapier workflows

Features 🚀✨

• 🎞️ Works with video
• 🎯 Handles single or multiple YouTube URLs via Input
• 📝 Fetches transcripts/captions when available
• 📊 Outputs full text, metadata (views, likes, author), and raw transcript segments

Input Configuration 🛠️

Provide input as a JSON object:

• startUrls (array): List of YouTube URLs to process.
• videoIds (array): List of YouTube Video IDs to process.
• lang (string, optional): Preferred language for transcripts (e.g., "en", "es").

Example Input

{
  "startUrls": [
    { "url": "https://www.youtube.com/watch?v=BthfXVCRWEQ" }
  ],
  "lang": "en"
}

Output 📦

For each video, the Actor pushes a result to the dataset containing:

• video_id: The ID of the video
• url: The input URL
• title: Video title
• description: Video description
• views: View count
• likes: Like count
• author: Channel information
• full_text: The complete transcript text joined together (Perfect for LLM context!)
• segments: Array of transcript segments with timestamps

Example Output

{
  "video_id": "BthfXVCRWEQ",
  "url": "https://www.youtube.com/watch?v=BthfXVCRWEQ",
  "title": "Hatching the World's Biggest Egg!",
  "description": "I'm hatching 3 eggs to see which one makes the best pet.",
  "views": 15000000,
  "likes": 500000,
  "author": {
    "name": "Mark Rober",
    "channelId": "UC...",
    "url": "..."
  },
  "duration": 600,
  "is_family_safe": true,
  "full_text": "Here is the world's biggest egg... and here is one of the smallest. I'm hatching 3 to see which one makes the best pet. ...",
  "segments": [
    {
      "start": 0.12,
      "end": 5.2,
      "text": "Here is the world's biggest egg...\nand here is one of the smallest."
    },
    {
      "start": 5.72,
      "end": 9.6,
      "text": "I'm hatching 3 to see which one makes the\nbest pet."
    },
    ...
  ]
}