Skip to content

Latest commit

 

History

History
35 lines (24 loc) · 1.23 KB

README.md

File metadata and controls

35 lines (24 loc) · 1.23 KB

AI-Powered Video Transcript Summarizer with Timestamps

This project automates the process of extracting subtitles from YouTube videos, formatting them with timestamps, and generating concise AI-powered summaries. It leverages advanced natural language processing models like ChatGPT to deliver meaningful insights into video content without requiring a full watch-through.

Features

  • Extracts subtitles (manual or auto-generated) from YouTube videos.
  • Formats subtitles with accurate timestamps in [hh:mm:ss - hh:mm:ss] format.
  • Summarizes the key points in the video while preserving timestamps.
  • Supports multiple languages (e.g., English, auto-generated English).
  • Flexible AI integration with OpenAI’s GPT models for summarization.

Tech Stack

  • Python: Core language.
  • pytube: Extracts video details and metadata.
  • youtube-transcript-api: Fetches video subtitles.
  • OpenAI API: For AI-driven summarization (using GPT-3.5 or GPT-4).
  • Transformers: Optional, for local summarization using Hugging Face models.

Installation

Prerequisites

  1. Python 3.8+
  2. Required Python libraries: 
    pip install pytube youtube-transcript-api openai
  3. Don't forget about OpenAI API key