YouTube Transcript Scraper

Retrieve complete transcripts for any YouTube video. Provide video or channel URLs, set data limits, and choose between timeline-based or continuous text output. You only pay for the data you scrape.

Overview

This actor scrapes the full transcript (captions) from YouTube videos. Input can be individual video URLs or channel URLs for bulk scraping. Output includes the transcript text, optional timestamps, and video metadata.

Key Features

• Target Flexibility: Scrape transcripts from specific video URLs or from multiple videos on a channel.
• Complete Transcripts: Retrieves the entire available caption set for each video.
• Output Customization: Choose your preferred transcript language and output format (timeline with segments or continuous text).
• Granular Billing: Pay per data element (video transcript) scraped.
• Simple Setup: Configure with a few input settings in the Apify Console.

How to Use

Configure the actor's input in the Apify Console.

1. Enter Target URLs

Provide a list of YouTube URLs. You can mix video and channel URLs in a single run.

2. Set a Video Crawl Limit (for channels)

If using channel URLs, set maxChannelVideos to limit how many recent videos to scrape (e.g., 5). This setting is ignored for direct video URLs. A value of 0 for a channel will result in no videos being scraped.

3. Configure Transcript Preferences

• transcriptFormat: Choose timeline (default) for segmented text with timestamps, or continuous for plain text.
• transcriptLanguage: Specify the language code (e.g., en) for the transcript if multiple are available.

Input

The actor accepts the following input configuration (typically as a JSON object in the Apify Console):

{
  "startUrls": [
    { "url": "https://www.youtube.com/watch?v=abc123" },
    { "url": "https://www.youtube.com/@ExampleChannel" }
  ],
  "maxChannelVideos": 10,
  "transcriptFormat": "timeline",
  "transcriptLanguage": "en"
}

• startUrls: (Required) Array of URLs to scrape.
• maxChannelVideos: Maximum number of videos to scrape from each channel URL. Required for channels.
• transcriptFormat: Output format. Options: "timeline", "continuous".
• transcriptLanguage: Preferred transcript language code (e.g., "en", "es").

Output

The actor outputs a dataset where each item represents one scraped video transcript. Each item includes:

• Video metadata (URL, title, ID, channel).
• The full transcript in the requested format.
• For timeline format: an array of segments with text, start (seconds), and duration (seconds).
• For continuous format: a single string containing the full transcript text.

Sample output item (timeline format):

{
  "url": "https://www.youtube.com/watch?v=abc123",
  "title": "Example Video Title",
  "transcript": [
    { "text": "First segment of speech.", "start": 0.0, "duration": 5.4 },
    { "text": "Next segment of speech.", "start": 5.4, "duration": 4.1 }
  ]
}