YouTube Transcript API

A FastAPI-based project for fetching YouTube video transcripts and processing them with OpenAI's GPT models for summarization and analysis.

---

Features

• Fetch YouTube video transcripts using the youtube-transcript-api.
• Summarize and analyze transcripts using OpenAI's GPT models.
• Expose APIs for transcript retrieval and optional processing.

---

Requirements

• Python 3.12+
• pip (Python package manager)

---

Setup Instructions

1. Clone the Repository

git clone https://github.com/your-repo/youtube-transcript-api.git
cd youtube-transcript-api

2. Create a Virtual Environment

python -m venv venv
source venv/bin/activate       # On Linux/Mac
.\venv\Scripts\activate        # On Windows

3. Install Dependencies

pip install -r requirements.txt

4. Configure Environment Variables

Create a .env file in the root directory:

OPENAI_API_KEY=your_openai_api_key_here

Replace your_openai_api_key_here with your actual OpenAI API key.

---

Run the Application

0. Prepare the Infrastructure

Ensure you have Docker and Docker Compose installed on your machine. Then, run the following command to set up the necessary infrastructure:

docker-compose up -d

This will start the required services defined in your docker-compose.yml file.

Apply Database Migrations

Run the following command to apply database migrations using yoyo-migrations:

yoyo apply

Replace username, password, and youtube_transcripts with your actual database credentials.

1. Start the Server

Run the FastAPI application using Uvicorn:

uvicorn app.__init__:app --reload

2. Access the API

• Swagger UI: http://127.0.0.1:8000/docs
• ReDoc: http://127.0.0.1:8000/redoc

---

Usage

Endpoints

1. Fetch Transcript

   GET /transcript/{video_id}
   

   • Parameters:
     • video_id (str): The YouTube video ID.
   • Response:

     {
       "video_id": "example_id",
       "transcript": "Transcript content..."
     }

2. Fetch and Summarize Transcript

   GET /transcript/{video_id}?summarize=true
   

   • Query Parameter:
     • summarize (bool): Set to true to include a summary.
   • Response:

     {
       "video_id": "example_id",
       "transcript": "Transcript content...",
       "summary": "Summarized content..."
     }

you can find more endpoint in the docs using swagger as explained later in this readme.

Development

Project Structure

youtube-transcript-api/
│
├── app/
│   ├── __init__.py            # FastAPI app entry point
│   ├── api/                   # API routes
│   │   ├── endpoints/
│   │   │   ├── transcript.py  # Transcript-related endpoints
│   ├── core/                  # Core configurations
│   │   ├── config.py          # Config file for environment variables
│   ├── services/              # Business logic
│   │   ├── openai_service.py  # Integration with OpenAI API
│   │   ├── transcript_service.py  # Logic for transcript fetching
│   ├── utils/                 # Utility functions
│
├── tests/                     # Tests for your application
├── .env                       # Environment variables
├── requirements.txt           # Dependencies
├── README.md                  # Project documentation

Run Tests

pytest

---

Contributing

Contributions are welcome! Feel free to open issues or submit pull requests.

---

License

This project is licensed under the MIT License.

---

Author

Josep Servat
For any queries, contact at DM on x @servatj or linkedin https://www.linkedin.com/in/servatj/.