This is a Streamlit-based web application that allows users to extract transcripts from YouTube videos and refine them using Google's Gemini AI. The app fetches the transcript, corrects errors in grammar, spelling, and punctuation, and improves readability while maintaining the original meaning of the video content.
- Extracts transcript from YouTube videos using the YouTube Transcript API.
- Uses Google Gemini AI to refine the transcript for better readability and clarity.
- Provides both the original and refined transcripts for comparison.
- Simple, user-friendly interface built with Streamlit.
Before you begin, ensure you have the following installed:
- Python 3.x
- Streamlit
- YouTube Transcript API
- Google Generative AI SDK (for accessing Gemini AI)
You can install the dependencies via pip:
pip install streamlit youtube-transcript-api google-generativeai- Clone this repository to your local machine:
git clone https://github.com/your-username/youtube-transcript-generator.git- Navigate to the project directory:
cd youtube-transcript-generator-
Create an environment variable for your Google API key or enter it directly in the app.
-
Run the Streamlit app:
streamlit run app.py- Open your browser to
http://localhost:8501to interact with the app.
- Input a valid YouTube video URL.
- Provide your Google API key for AI processing.
- The app will display the original and refined transcripts.
To use the Gemini AI model, you need to generate an API key from Google Cloud:
- Visit the Google Cloud Console.
- Enable the Google Generative AI API.
- Generate an API key and provide it to the app.
├── youtube.py # Main Streamlit app file
├── README.md # This file
└── requirements.txt # Required packages for the project
This project is licensed under the MIT License. See the LICENSE file for details.
For any queries or further information, feel free to reach out via LinkedIn or email at mr.nithesh.k@gmail.com.