YouTube Transcript

Extract transcripts and captions from YouTube videos.

When to Use

Use this skill when the user wants to:

• Get the transcript of a YouTube video
• Extract captions or subtitles
• Analyze video content without watching
• Create summaries or notes from videos
• Search for specific quotes in videos
• Convert video content to text format

Installation

npm install youtube-transcript

Or use Python:

pip install youtube-transcript-api

JavaScript/Node.js Usage

const { YoutubeTranscript } = require('youtube-transcript');

// Extract transcript
const transcript = await YoutubeTranscript.fetchTranscript('VIDEO_ID');

// transcript is an array of objects:
// [
//   { text: 'Hello', duration: 1.5, offset: 0 },
//   { text: 'World', duration: 2.0, offset: 1.5 },
//   ...
// ]

// Get full text
const fullText = transcript.map(item => item.text).join(' ');
console.log(fullText);

Python Usage

from youtube_transcript_api import YouTubeTranscriptApi

# Get transcript
transcript = YouTubeTranscriptApi.get_transcript('VIDEO_ID')

# With language preference
transcript = YouTubeTranscriptApi.get_transcript('VIDEO_ID', languages=['en', 'fr'])

# Full text
full_text = ' '.join([item['text'] for item in transcript])
print(full_text)

Extract Video ID from URL

function getVideoId(url) {
  const match = url.match(/(?:v=|\/)([\w-]{11})(?:\?|&|$)/);
  return match ? match[1] : null;
}

// Examples:
// https://www.youtube.com/watch?v=dQw4w9WgXcQ → dQw4w9WgXcQ
// https://youtu.be/dQw4w9WgXcQ → dQw4w9WgXcQ

Common Patterns

Save transcript to file

const fs = require('fs');
const { YoutubeTranscript } = require('youtube-transcript');

async function saveTranscript(videoId, outputFile) {
  const transcript = await YoutubeTranscript.fetchTranscript(videoId);
  const text = transcript.map(item => item.text).join(' ');
  fs.writeFileSync(outputFile, text);
}

await saveTranscript('dQw4w9WgXcQ', 'transcript.txt');

Formatted transcript with timestamps

async function getFormattedTranscript(videoId) {
  const transcript = await YoutubeTranscript.fetchTranscript(videoId);

  return transcript.map(item => {
    const time = new Date(item.offset * 1000).toISOString().substr(11, 8);
    return `[${time}] ${item.text}`;
  }).join('\n');
}

console.log(await getFormattedTranscript('VIDEO_ID'));
// Output:
// [00:00:00] Hello
// [00:00:01] World

Search in transcript

async function searchTranscript(videoId, query) {
  const transcript = await YoutubeTranscript.fetchTranscript(videoId);

  return transcript
    .filter(item => item.text.toLowerCase().includes(query.toLowerCase()))
    .map(item => ({
      time: item.offset,
      text: item.text,
      url: `https://youtube.com/watch?v=${videoId}&t=${Math.floor(item.offset)}s`
    }));
}

const results = await searchTranscript('VIDEO_ID', 'important');

Multiple Languages

from youtube_transcript_api import YouTubeTranscriptApi

# List available transcripts
transcript_list = YouTubeTranscriptApi.list_transcripts('VIDEO_ID')

for transcript in transcript_list:
    print(f"Language: {transcript.language}")
    print(f"Generated: {transcript.is_generated}")

# Get in specific language
transcript = YouTubeTranscriptApi.get_transcript('VIDEO_ID', languages=['fr'])

Batch Processing

async function getMultipleTranscripts(videoIds) {
  const results = [];

  for (const id of videoIds) {
    try {
      const transcript = await YoutubeTranscript.fetchTranscript(id);
      results.push({ id, transcript, error: null });
    } catch (error) {
      results.push({ id, transcript: null, error: error.message });
    }
  }

  return results;
}

Common Issues

Transcript not available: Some videos don't have captions

try {
  const transcript = await YoutubeTranscript.fetchTranscript(videoId);
} catch (error) {
  console.log('No transcript available for this video');
}

Rate limiting: Add delays between requests

await new Promise(resolve => setTimeout(resolve, 1000)); // Wait 1s

Language not found: List available languages first

transcript_list = YouTubeTranscriptApi.list_transcripts('VIDEO_ID')
for t in transcript_list:
    print(t.language_code)

Best Practices

• Always handle errors (not all videos have transcripts)
• Cache transcripts to avoid repeated API calls
• Respect YouTube's terms of service
• Add delays for batch processing
• Check transcript availability before processing
• Use generated transcripts as fallback

Resources

• Node.js: https://www.npmjs.com/package/youtube-transcript
• Python: https://github.com/jdepoix/youtube-transcript-api
• YouTube Data API: https://developers.google.com/youtube/v3