Subtitle Generator

AI-powered subtitle generation using Whisper for accurate speech-to-text transcription.

Key Features:

🎯 Multi-format output - VTT, SRT, TXT, JSON, LRC, ASS, TTML
🚀 Fast processing - Powered by whisper.cpp for high-performance inference
📦 Batch processing - Process multiple videos at once
🔄 Video embedding - Embed subtitles directly into videos
🌍 Multilingual - Support for multiple languages
🔓 Open-source - Freely available for use, modification, and distribution

Installation

Quick Install (PyPI)

pip install subtitle-generator

Note: FFmpeg is required. Install via: brew install ffmpeg (macOS) or sudo apt install ffmpeg (Ubuntu)

Development Setup

For contributors or if you need to build whisper.cpp from source:

Prerequisites

git, make, cmake
ffmpeg (Required for video processing)
conda (Anaconda or Miniconda)

Setup

Clone and setup Whisper.cpp:
```
./setup_whisper.sh
```

Create and activate conda environment:

conda env create -f environment.yml
conda activate subtitle

Usage

Generate Subtitles

# Basic usage (generates VTT subtitle file)
python subtitle.py video.mp4

# Generate and embed subtitles into video
python subtitle.py video.mp4 --merge

# Use a specific model
python subtitle.py video.mp4 --model base

# Generate SRT format
python subtitle.py video.mp4 --format srt

# From URL
python subtitle.py "https://example.com/video.mp4"

Model Management

# List all available models
python subtitle.py models --list

# Download a specific model
python subtitle.py models --download large

View Supported Formats

python subtitle.py formats

Options

| Option | Description | |--------|-------------| | --model, -m | Model to use (default: base) | | --format, -f | Output format: vtt, srt, txt, json, lrc (default: vtt) | | --merge | Embed subtitles into video | | --threads, -t | Number of threads (default: 4) | | --verbose, -v | Verbose output |

Available Models

| Model | Size | Speed | Best For | |-------|------|-------|----------| | tiny | ~75MB | Fastest | Quick previews | | base | ~140MB | Fast | General use (default) | | small | ~460MB | Medium | Quality output | | medium | ~1.5GB | Slow | Professional work | | large | ~3GB | Slowest | Maximum accuracy |

Tip: Use .en models (e.g., base.en) for English-only content.

Documentation

Usage Examples - CLI and Python API examples
API Reference - Programmatic API documentation
Troubleshooting - Common issues and solutions
Contributing - Contribution guidelines

License

MIT

Reference & Credits

Author

Ved Gupta

Support

For support, email vedgupta@protonmail.com

Subtitle

Install / Use

README