YouTube Video Info Extractor — Metadata & Transcripts

YouTube Video Info Extractor — Metadata & Transcripts

Extract detailed metadata from YouTube videos — title, description, channel info, publish date, duration, view/like/comment counts, tags, thumbnail URL, and optional transcripts/subtitles. Whether you're building a content analysis pipeline, creating a video archive, or researching YouTube trends, this actor extracts everything you need in a single call.

---

What does it do?

This actor takes one or more YouTube video URLs and extracts comprehensive metadata from each one using yt-dlp — the most reliable YouTube data extraction library. For each video you get the title, channel information, publishing date, duration (ISO 8601), engagement metrics (views, likes, comments), tags, thumbnail URL, and optionally the full transcript or subtitles.

The actor handles all the edge cases: private videos, deleted content, age-restricted uploads, and region-blocked videos are caught gracefully and reported without halting the batch. Results are streamed to the dataset as each video completes, so you can start consuming data before the full batch finishes.

Concurrent processing with a configurable semaphore (max 2 simultaneous extractions) balances speed against YouTube's rate limits, and the actor charges one video-extract event per video processed.

Features

1. Full metadata extraction — Title, channel name, channel URL, published date, ISO 8601 duration, view count, like count, comment count, tags, and thumbnail URL.
2. Optional transcript support — Fetch video transcripts/subtitles as plain text when includeTranscript is enabled. Automatically picks English subtitles (manual or auto-generated) with fallback to any available language.
3. Batch processing — Submit multiple YouTube URLs in a single run. Process up to 50 videos at once (configurable via maxResults).
4. Concurrent extraction — Up to 2 simultaneous yt-dlp extractions via asyncio semaphore, balancing throughput against rate limits.
5. Graceful error handling — Private videos, deleted content, age-restricted uploads, region blocks, and copyright takedowns are reported per-video without affecting other results in the batch.
6. ISO 8601 durations — Video duration is returned in standard ISO 8601 format (e.g., PT5M30S, PT1H2M15S) for easy parsing.
7. ISO 8601 timestamps — Publish dates are returned in ISO 8601 format (YYYY-MM-DDTHH:MM:SSZ).

Why use this?

Problem	Solution
You need to analyze a large set of YouTube videos programmatically	Submit all URLs in one batch and get structured metadata for every video
You want to build a searchable video archive	Extract tags, descriptions, and transcripts for full-text indexing
You're researching content trends or channel performance	Get view counts, like counts, and comment counts for every video
You need clean transcript text for NLP or summarization	Enable transcript extraction once and get plain-text output
You're migrating or backing up YouTube content metadata	Batch-extract all metadata in a structured, machine-readable format
You want to know why a video failed extraction	Every error is classified (private, deleted, age-restricted, etc.) per video

Who is it for?

Persona	What they use it for
Content Analyst	Extracting metadata and transcripts from video playlists for trend analysis
Data Engineer	Building ETL pipelines that ingest YouTube metadata into data warehouses
Researcher	Collecting video metadata and transcripts for academic NLP or social science research
Creator / Marketer	Auditing a competitor's video library for SEO tags, engagement stats, and content patterns
Archive Curator	Cataloguing video metadata before taking down or migrating content
Developer	Building applications that need programmatic access to YouTube video details

Input Parameters

Field	Type	Default	Description
urls	array	—	Array of YouTube video URLs to process (e.g., ["https://www.youtube.com/watch?v=..."])
includeTranscript	boolean	false	Whether to fetch and include video transcript/subtitles (English preferred, auto-generated or manual)
maxResults	integer	10	Maximum number of videos to process. Range: 1-50.

Example Input JSON

{
  "urls": ["https://www.youtube.com/watch?v=dQw4w9WgXcQ"],
  "includeTranscript": false,
  "maxResults": 10
}

Example Input with Transcript

{
  "urls": [
    "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
    "https://www.youtube.com/watch?v=jNQXAC9IVRw"
  ],
  "includeTranscript": true,
  "maxResults": 5
}

Output Format

Field	Type	Description
url	string	The YouTube video URL that was processed
title	string or null	Title of the YouTube video
channel	string or null	Name of the YouTube channel that published the video
channelUrl	string or null	URL to the YouTube channel
publishedAt	string or null	ISO 8601 date when the video was published (YYYY-MM-DDTHH:MM:SSZ)
duration	string or null	Video duration in ISO 8601 format (e.g., PT5M30S, PT1H2M15S)
viewCount	integer or null	Number of views
likeCount	integer or null	Number of likes
commentCount	integer or null	Number of comments
tags	array	Array of video tags/keywords
thumbnailUrl	string or null	URL to the video thumbnail image
transcript	string or null	Video transcript/subtitles as plain text (null if not requested or unavailable)
error	string or null	Error message if extraction failed (null on success)

Example Output JSON

{
  "url": "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
  "title": "Rick Astley - Never Gonna Give You Up (Official Music Video)",
  "channel": "Rick Astley",
  "channelUrl": "https://www.youtube.com/channel/UCuAXFkgsw1L7xaCfnd5JJOw",
  "publishedAt": "2009-10-25T00:00:00Z",
  "duration": "PT3M33S",
  "viewCount": 1500000000,
  "likeCount": 20000000,
  "commentCount": 5000000,
  "tags": ["Rick Astley", "Never Gonna Give You Up", "80s", "pop", "music video"],
  "thumbnailUrl": "https://i.ytimg.com/vi/dQw4w9WgXcQ/maxresdefault.jpg",
  "transcript": null,
  "error": null
}

Example Error Output

{
  "url": "https://www.youtube.com/watch?v=INVALID_ID",
  "title": null,
  "channel": null,
  "channelUrl": null,
  "publishedAt": null,
  "duration": null,
  "viewCount": null,
  "likeCount": null,
  "commentCount": null,
  "tags": [],
  "thumbnailUrl": null,
  "transcript": null,
  "error": "This video is unavailable"
}

API Usage

cURL

curl -X POST "https://api.apify.com/v2/acts/perryay~youtube-video-info-extractor/runs?token=YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "urls": ["https://www.youtube.com/watch?v=dQw4w9WgXcQ"],
    "includeTranscript": false,
    "maxResults": 10
  }'

Python (ApifyClient)

from apify_client import ApifyClient

client = ApifyClient("YOUR_API_TOKEN")
run = client.actor("perryay~youtube-video-info-extractor").call(
    run_input={
        "urls": ["https://www.youtube.com/watch?v=dQw4w9WgXcQ"],
        "includeTranscript": False,
        "maxResults": 10
    }
)
dataset = client.dataset(run["defaultDatasetId"]).list_items()
for item in dataset.items:
    title = item.get("title", "N/A")
    channel = item.get("channel", "N/A")
    views = item.get("viewCount", "N/A")
    error = item.get("error")
    if error:
        print(f"[ERROR] {item['url']}: {error}")
    else:
        print(f"{title} — {channel} — {views} views")

Use Cases

1. Content research and trend analysis — Extract metadata from hundreds of videos to analyze what drives engagement. Compare view counts, like ratios, and comment volumes across topics, channels, or time periods.

2. Video archive and backup — Before taking down or restructuring video content, batch-extract all metadata including tags, descriptions, and timestamps for your records.

3. NLP and transcript analysis — Enable transcript extraction to get clean text from thousands of videos. Use the text for summarization, keyword extraction, sentiment analysis, or training custom ML models.

4. Competitive channel audit — Extract metadata from a competitor's video library to understand their content strategy — which tags they use, how long their videos are, and what engagement looks like.

5. SEO and tag research — Collect tags from top-performing videos in your niche. Identify common keywords and patterns to optimize your own video metadata for discoverability.

6. Playlist and channel migration — Moving content between YouTube accounts or platforms? Extract all metadata in advance to preserve titles, descriptions, and tags during the transition.

7. Content moderation monitoring — Periodically re-check video metadata to detect changes in titles, descriptions, or availability. Flag videos that have been made private or had their metadata altered.

8. Dataset creation for ML — Build training datasets for video recommendation systems, content classification models, or engagement prediction by extracting structured metadata at scale.

9. Creator portfolio analysis — For content creators, analyze your own video library to see which tags, titles, and durations correlate with higher engagement. Make data-driven content decisions.

10. Integration with data pipelines — Combine with other Apify actors (Domain Intel, Link Quality Analyzer) to build comprehensive media monitoring and analysis pipelines.

FAQ

Q: What video information can this actor extract? A: It extracts title, channel name, channel URL, published date (ISO 8601), duration (ISO 8601), view count, like count, comment count, tags, thumbnail URL, and optionally the transcript or subtitles in plain text.

Q: Does the actor download the actual video file? A: No. The actor only extracts metadata and optionally fetches subtitle text. No video or audio content is downloaded.

Q: How many URLs can I submit at once? A: The maxResults parameter caps the number of videos processed per run (default 10, max 50). The urls array can contain more entries, but only the first maxResults will be processed.

Q: How long does it take to process a video? A: Each video takes roughly 2-10 seconds depending on YouTube's response time and network conditions. The actor processes up to 2 videos concurrently. For a batch of 10 videos, expect 15-60 seconds total.

Q: How do transcripts work? A: When includeTranscript is true, the actor fetches the best available English subtitle track (manual uploads preferred, auto-generated captions as fallback). If English subtitles are unavailable, it falls back to any available language. The subtitles are converted to clean plain text.

Q: What happens if a video doesn't have subtitles? A: The transcript field will be null when no subtitles or captions are available for the video. This is not treated as an error — the rest of the metadata is returned normally.

Q: Can the actor extract info from age-restricted videos? A: Age-restricted videos return an error message: "This video is age-restricted and cannot be accessed without authentication." Other videos in the batch are unaffected.

Q: What about private, deleted, or unavailable videos? A: Each of these scenarios is detected and reported in the error field with a human-readable message. The actor continues processing the remaining URLs in the batch.

Q: Does it work with YouTube Shorts? A: Yes. YouTube Shorts URLs (e.g., https://www.youtube.com/shorts/...) are supported and return the same metadata structure as regular videos.

Q: Can I extract metadata from a YouTube playlist? A: This actor processes individual video URLs, not playlists. To extract metadata from an entire playlist, provide each video URL in the urls array. For playlist-level extraction, consider dedicated YouTube playlist tools.

Q: Does the actor require a YouTube API key? A: No. The actor uses yt-dlp, which extracts publicly available metadata without requiring API keys or OAuth authentication.

Q: What error types are handled? A: The actor handles: private videos, deleted/removed videos, age-restricted content, copyright-blocked videos, region-unavailable content, members-only videos, sign-in-required content, and general network/parsing errors. Each is returned with a descriptive error message per video.

Q: What's the concurrency limit and why? A: The actor limits simultaneous extractions to 2 to stay within YouTube's comfortable rate limit boundaries. Higher concurrency would risk rate limiting and would not meaningfully improve throughput.

Q: Can I run this actor on a schedule? A: Yes. The actor has a fully documented API and can be integrated into any automation workflow, scheduler, or CI/CD pipeline via the Apify API.

Usage & Billing

This actor uses Apify's PAY_PER_EVENT pricing model. You are charged per event:

Event Name	Price (USD)	Trigger
apify-actor-start	$0.020	Charged on every actor start
video-extract	$0.010	Charged per video processed
transcript-fetch	$0.010	Charged when transcript is fetched
batch-extract	$0.005	Charged per batch run (multiple videos)

Platform infrastructure costs (Apify's compute and storage) are passed through at cost.

MCP Integration

This actor can be used through the Apify MCP server. Once connected, your MCP client (Claude Desktop, Cursor, etc.) can discover and run this actor from the Apify Store.

Quick Start

1. Install the Apify connector in your MCP client:

   • Claude Desktop: Search for "Apify" in the connector directory, or use the remote server at https://mcp.apify.com
   • Other clients: See the Apify MCP server docs for setup instructions

2. Ask your AI assistant to use the actor.

Claude Desktop Configuration

Add the following to your claude_desktop_config.json:

{
  "mcpServers": {
    "apify": {
      "url": "https://mcp.apify.com"
    }
  }
}

On first connection, your browser opens to sign in to Apify and authorize access.

Bearer Token Alternative

For MCP clients that do not support browser-based OAuth (e.g., Cursor, VS Code, Windsurf), use a Bearer token directly:

{
  "mcpServers": {
    "apify": {
      "url": "https://mcp.apify.com",
      "headers": {
        "Authorization": "Bearer <APIFY_TOKEN>"
      }
    }
  }
}

Replace <APIFY_TOKEN> with your actual Apify API token from Apify Console.

Related Tools

• Meta Mate — Extract Open Graph, Twitter Cards, and JSON-LD metadata from URLs
• Link Quality Analyzer — Deep analysis of links including HTTP status, redirect chains, and security posture
• Domain Intel — WHOIS lookups, DNS enumeration, and SSL certificate validation for brand protection and domain research
• UUID Lab — Generate and decode UUID v4, v7, and nanoid identifiers for consistent cross-platform user and session tracking
• Social Media Username Checker — Check username availability across 50+ social media platforms
• JSON Studio — Format, validate, diff, transform, and explore JSON documents
• QR Craft — Generate high-quality QR codes in PNG or SVG format