YouTube Video Summarizer - Fixed Version 🔧 This Apify actor extracts video metadata and generates AI-powered summaries from YouTube videos. ## 🚨 Recent Fixes & Features (October 2025) ### Issues Resolved: 1. ✅ Bot Detection Bypass: Enhanced yt-dlp configuration with proper headers and user agents 2. ✅ Proxy Connection Timeouts: Improved proxy handling with intelligent fallback mechanisms 3. ✅ Multiple Extraction Strategies: Implemented 5 different extraction approaches 4. ✅ Better Error Handling: Clean logging with only relevant information 5. ✅ Azure OpenAI Integration: Pre-configured with your Azure OpenAI credentials 6. ✅ Smart Proxy Management: Follows Instagram scraper pattern with proper fallbacks ### Key Improvements: - 🤖 Azure OpenAI Pre-configured: Ready to generate summaries out of the box - 🔄 Smart Proxy Fallback: User proxy → Residential proxy (3 retries) → Direct - 🧪 Local Testing Support: Uses hardcoded residential proxy for development - 📊 Clean Logging: Suppressed verbose yt-dlp errors, shows only relevant info - 🚀 Production Ready: Cleaned up all test files, optimized for Apify deployment ## Features - Streams results live to the Dataset (one per URL) - Proxy: DIRECT → USER (if provided) → RESIDENTIAL fallback on rejection; then sticks to it - Retries + backoff; consistent output (status + error fields) - Set OPENAI_API_KEY to enable summaries - Robust error handling and retry mechanisms ## 🚀 Deployment Instructions ### For Apify Platform: 1. Upload the entire youtube-video-summerizer folder to your Apify actor 2. Ensure the Dockerfile and requirements.txt are properly configured 3. Set up proxy configuration in the actor input (recommended: use Apify Proxy with RESIDENTIAL group) 4. Add your OpenAI API key to the actor environment variables ### Input Configuration: json { "urls": ["https://www.youtube.com/watch?v=VIDEO_ID"], "maxRetries": 3, "proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"] } } ### Environment Variables: ✅ Azure OpenAI (Pre-configured): - The actor comes with Azure OpenAI pre-configured and ready to use - Default endpoint: https://amzautopilot002.openai.azure.com - Default deployment: gpt-4o-4 - API version: 2023-12-01-preview Override Options (Optional): - AZURE_OPENAI_ENDPOINT: Override the default Azure OpenAI endpoint - AZURE_OPENAI_API_KEY: Override the default Azure OpenAI API key - AZURE_OPENAI_DEPLOYMENT: Override the default deployment name - AZURE_OPENAI_API_VERSION: Override the default API version - OPENAI_API_KEY: Use regular OpenAI instead of Azure OpenAI Platform Variables: - APIFY_PROXY_PASSWORD: Automatically provided by Apify platform ## 🚀 Deployment Ready for Apify Platform: 1. Upload this entire folder to your Apify actor 2. The actor is pre-configured with Azure OpenAI credentials 3. Set up proxy configuration in the input (recommended: RESIDENTIAL) 4. Deploy and run! Local Testing: - Automatically uses hardcoded residential proxy - All Azure OpenAI features work out of the box - Clean logging shows only relevant information ## 📝 Error Handling The actor now provides more specific error messages: - Bot Detection: "YouTube detected automated access. Please try again later or use different proxy settings." - Connection Issues: "Connection timeout occurred. Please check your internet connection or proxy settings." - Restricted Content: "Video is private, unavailable, or restricted." - Age Restrictions: "Video is age-restricted and cannot be accessed." ## 🔄 Retry Logic The actor implements intelligent retry logic: 1. Direct Connection: First attempt without proxy 2. User Proxy: If configured, switches to user-specified proxy 3. Residential Fallback: Automatically falls back to residential proxy on failures 4. Multiple Strategies: Tries 5 different yt-dlp configurations 5. Smart Delays: Implements exponential backoff with connection-aware delays Build & Run - .actor/actor.json → "dockerfile": "../Dockerfile", "input": "./input_schema.json" - apify run locally, or Deploy & Run on Apify