Youtube video summarizer

This Apify actor extracts subtitles from YouTube videos and generates AI-powered summaries using OpenAI's API. Perfect for quickly understanding long video content, creating video summaries for research, or content curation.

Features

• 🎥 Extract subtitles from any YouTube video with available captions
• 🤖 Generate intelligent summaries using OpenAI's GPT models
• 🌍 Support for translated subtitles in different languages
• ⚙️ Configurable AI parameters (model, temperature, max tokens)
• 📊 Detailed output including metadata and usage statistics
• 🔒 Secure handling of API keys

How it Works

1. Input: Provide a YouTube video URL and your API key
2. Fetch Subtitles: Retrieves subtitles
3. AI Summary: Sends the transcript to OpenAI for intelligent summarization
4. Output: Returns a structured summary with metadata

Input Parameters

Required Fields

• OpenAI API Key: Your OpenAI API key for generating summaries
• Video URL: YouTube video full URL

Optional Configuration

• OpenAI Model: Select from the latest models:
  • GPT-4o Mini: Fastest and most cost-effective (16K output)
  • GPT-4o: Balanced performance and cost (16K output)
  • GPT-4.1 Nano: Ultra-fast with 1M context window (32K output)
  • GPT-4.1 Mini: Best value with 1M context window (32K output)
  • GPT-4.1: Most capable with 1M context window (32K output)
  • o4-Mini: Advanced reasoning model (100K output, 200K context)
• Max Tokens: Control summary length (100-4000 tokens)
• Temperature: Adjust AI creativity (0-1, where 0 is more focused)
• Custom Prompt: Provide specific instructions for summarization
• Include Full Transcript: Option to include complete transcript in output

Input Example

{
  "videoUrl": "https://www.youtube.com/watch?v=ycR4s5VOlng",
  "openaiApiKey": "your-openai-api-key-here",
  "openaiModel": "gpt-3.5-turbo",
  "maxTokens": 800,
  "temperature": 0.7,
  "includeFullTranscript": false
}

Output Format

The actor outputs a JSON object with the following structure:

{
  "videoId": "ycR4s5VOlng",
  "videoUrl": "https://www.youtube.com/watch?v=ycR4s5VOlng",
  "summary": "AI-generated summary of the video content...",
  "transcriptLength": 1250,
  "subtitleSegments": 45,
  "timestamp": "2024-01-15T10:30:00.000Z",
  "openaiModel": "gpt-3.5-turbo",
  "tokensUsed": 650,
  "fullTranscript": "Complete transcript..." // if includeFullTranscript is true
}

API Requirements

OpenAI API Access

1. Create an account at OpenAI
2. Generate an API key in your dashboard
3. Ensure you have sufficient credits for API usage

Use Cases

• Content Research: Quickly understand long educational videos
• Meeting Summaries: Summarize recorded video meetings
• Content Curation: Generate descriptions for video libraries
• Academic Research: Extract key points from lecture videos
• Social Media: Create summaries for video sharing
• Accessibility: Provide text summaries for audio-visual content

Custom Prompts

You can customize the summarization by providing a custom prompt. Use {transcript} as a placeholder for the video transcript:

"Please create a bullet-point summary of this video focusing on:
- Main topics discussed
- Key insights and takeaways
- Any actionable recommendations

Transcript: {transcript}"

Error Handling

The actor handles various error scenarios:

• Invalid Video: Returns error if video ID/URL is invalid
• No Subtitles: Fails gracefully if video has no available subtitles
• API Limits: Provides clear messages for rate limiting or quota issues
• Authentication: Validates API keys and provides helpful error messages

Performance Notes

• Processing time depends on video length and OpenAI model choice
• GPT-4o Mini/4.1 Nano: Fastest response times (1-10 seconds)
• GPT-4.1 Mini: Good balance of speed and quality (5-20 seconds)
• GPT-4.1/o4-Mini: Highest quality but slower (10-60 seconds)
• Large Context Models: GPT-4.1 series can handle very long videos with 1M token context
• Subtitle extraction is typically very fast (1-5 seconds)
• Model selection automatically enforces appropriate token limits

Limitations

• Only works with videos that have available subtitles/captions
• Summary quality depends on subtitle accuracy
• Subject to OpenAI rate limits and pricing
• Some videos may have restricted subtitle access

Cost Considerations

• OpenAI: Varies by model and content length (pricing as of 2025)
  • GPT-4o Mini: $0.15/M input, $0.60/M output tokens
  • GPT-4o: $2.50/M input, $10.00/M output tokens
  • GPT-4.1 Nano: $0.10/M input, $0.40/M output tokens
  • GPT-4.1 Mini: $1.00/M input, $4.00/M output tokens
  • GPT-4.1: $3.00/M input, $12.00/M output tokens
  • o4-Mini: $3.00/M input, $12.00/M output tokens
• Average cost per video summary: $0.01-0.15 depending on model and video length

Support

For issues or questions:

1. Check the actor logs for detailed error messages
2. Verify your API keys are valid and have sufficient credits
3. Ensure the YouTube video has available subtitles
4. Contact support through the Apify platform