Audio/Video Transcription Service

A web application that transcribes audio and video files using OpenAI's Whisper API, with GPT-4o post-processing for improved formatting.

Preview

Combined with chinese translation chrome extension:

Features

• Audio and video file transcription
• Support for large files (auto-splitting)
• Markdown formatted output
• Real-time processing status
• Transcription history
• Copy to clipboard functionality
• Dark mode support
• Background task processing

Supported File Formats

• Audio: MP3, WAV, M4A, MPGA
• Video: MP4, WEBM

Prerequisites

System Requirements

• Python 3.8 or higher
• Redis server
• ffmpeg (for audio processing)

Installing System Dependencies

Ubuntu/Debian

bash

Install Redis

sudo apt-get update sudo apt-get install redis-server

Install ffmpeg

sudo apt-get install ffmpeg

configure openai api key

create a file named .env in the root directory and add the following line: OPENAI_API_KEY=your_api_key

1. Start Redis server (if not already running):

sudo service redis-server start # On Ubuntu/Debian

2. Start Celery worker (in a separate terminal):

celery -A celery_worker.celery worker --loglevel=info --concurrency=4

3. Run the Flask application:

python app.py