YouTube Thumbnail Generator

An AI-powered tool that generates YouTube thumbnails using multiple specialized agents working together. The system analyzes video content, processes images, and generates engaging thumbnails using OpenAI's gpt-image1 model.

Architecture

The system consists of several specialized agents:

1. Summary Agent (summaryAgent.py)

• Analyzes YouTube videos using the YouTube Transcript API
• Extracts video content and generates concise summaries
• Uses Google's Gemini AI for intelligent content analysis
• Handles various video formats and transcript availability

2. Classification Agent (classificationAgent.py)

• Analyzes input images using Google's Gemini Vision
• Identifies image types (person, logo, background, etc.)
• Provides detailed descriptions and usage suggestions
• Helps in determining how images can be used in thumbnails

3. Prompt Agent (promptAgent.py)

• Generates detailed prompts for image generation
• Combines video summary and image analysis
• Incorporates brand themes and styling preferences
• Creates optimized prompts for gpt-image-1

4. Image Generation Agent (imageGen.py)

• Handles image generation and editing using gpt-image-1
• Supports both text-to-image and image-to-image generation
• Manages image quality and size parameters
• Saves generated thumbnails with timestamps

5. Main Orchestrator (main.py)

• Coordinates all agents in a logical workflow
• Manages user input and agent interactions
• Handles the complete thumbnail generation process
• Provides a user-friendly command-line interface

Setup

1. Clone the repository
2. Install dependencies:

   pip install -r requirements.txt

3. Create a .env file with your API keys:

   OPENAI_API_KEY=your_openai_api_key_here
   GOOGLE_API_KEY=your_google_api_key_here
   

Usage

Run the main orchestrator:

python main.py

The tool will guide you through the following steps:

1. Video Analysis

   • Enter a YouTube URL
   • System analyzes video content and generates a summary

2. Image Processing

   • Choose to use existing images or generate new ones
   • If using existing images, provide image paths
   • System analyzes images and suggests usage

3. Prompt Generation

   • Optionally provide brand theme
   • System generates optimized prompt for thumbnail

4. Thumbnail Generation

   • System generates or edits images based on the prompt
   • Saves the final thumbnail with timestamp

Features

• Intelligent Video Analysis

  • Automatic transcript extraction
  • Smart content summarization
  • Fallback options for videos without transcripts

• Advanced Image Processing

  • Multi-image analysis
  • Intelligent image classification
  • Usage suggestions for thumbnail composition

• Smart Prompt Generation

  • Context-aware prompt creation
  • Brand theme integration
  • Style and composition suggestions

• Flexible Image Generation

  • Text-to-image generation
  • Image-to-image editing
  • High-quality output
  • Customizable parameters

Requirements

• Python 3.8+
• OpenAI API key
• Google API key
• Required Python packages (see requirements.txt)

Error Handling

The system includes robust error handling for:

• Invalid YouTube URLs
• Missing transcripts
• Image processing errors
• API limitations
• Network issues

Output

Generated thumbnails are saved as PNG files with timestamps in the format:

thumbnail_[timestamp].png