oneword-ai 0.1.1

Create viral-style one-word subtitles for Reels, Shorts & TikToks using AI. Perfect for content creators!

🎬 OneWord AI - Cinematic Subtitle Generator

Generate viral-style one-word subtitles from video/audio using Whisper AI

Perfect for creating high-energy reels, shorts, and TikToks! 🚀

Installation • Usage • Features • Credits

---

✨ Features

• 🎯 Three Subtitle Modes: One Word, Two Word Punch, Phrase Mode
• 🌍 Multi-Language Support: Auto-detect or specify (English, Hindi, Urdu, Spanish)
• 🤖 Multiple AI Models:
  • OpenAI Whisper (Medium, Large)
  • Hindi2Hinglish (Oriserve/Whisper-Hindi2Hinglish-Prime) 🇮🇳
• 💻 Dual Interface: CLI for power users, Web UI for visual workflow
• 📦 Easy Installation: One-line pip install
• ☁️ Cloud Ready: Works seamlessly on Google Colab

---

🚀 Installation

Method 1: Install from PyPI (Recommended)

pip install oneword-ai

Prerequisites: Ensure FFmpeg is installed on your system.

---

📖 Usage

Option 1: Web Interface (Easiest)

Start the web server and process files through a beautiful UI:

# Start server
python -m onewordai.api.main

Then open http://localhost:8000 in your browser.

Features:

• 📤 Drag & drop file upload
• 📊 Real-time download progress with speed tracking
• ⏱️ Live transcription status
• ❌ Cancel processing anytime
• 📥 Instant SRT download
• ⚠️ Reload protection (won't lose progress)

---

Option 2: Google Colab with Package

Run OneWord AI in the cloud without any local installation:

Quick Start in Colab:

# Install package
!pip install oneword-ai

# Import and use
from onewordai.core.engine import SubtitleGenerator

# Generate subtitles
generator = SubtitleGenerator(model_name="medium")
output = generator.process("video.mp4", mode="oneword")
print(f"✅ Subtitles saved: {output}")

---

Option 3: Command Line Interface (CLI)

For batch processing and automation:

# Basic usage
python -m onewordai.cli -i video.mp4

# Advanced options
python -m onewordai.cli -i video.mp4 -m medium -lang hi -mode oneword

See CLI.md for full CLI documentation.

---

📊 Subtitle Modes

Mode	Description	Best For
One Word	Each word = separate subtitle	High-energy viral content, reels
Two Word Punch	Groups of 2 words	Punchy messaging, Instagram posts
Phrase Mode	Full sentence segments	YouTube videos, longer content

---

🎨 Model Options

Model	Size	Speed	Quality	Language Support
medium	~1.5GB	Fast	Good	Multi-language
large	~3GB	Slower	Best	Multi-language
Hindi2Hinglish	~1.5GB	Fast	Excellent for Hindi	Hindi → Hinglish

---

🤝 Credits

This project wouldn't be possible without these amazing open-source projects:

Core Technologies

• OpenAI Whisper - Revolutionary speech recognition model
• Oriserve/Whisper-Hindi2Hinglish-Prime - Hindi to Hinglish transcription model
• HuggingFace Transformers - Model loading and inference

Backend & UI

• FastAPI - Modern Python web framework
• Gradio - ML web interfaces
• FFmpeg - Audio/video processing

Special Thanks

• OpenAI for making Whisper open-source
• Oriserve team for the Hindi2Hinglish model
• All contributors to the dependencies

---

📜 License

MIT License - see license.txt

Free to use for personal and commercial projects!

---

👨‍💻 Author

Built with ❤️ by Ambrish Yadav

💼 Connect: @ambrish.yadav.1

---

⭐ Star this repo if you find it useful! ⭐

Report Bug • Request Feature