Automatically add any language subtitles to your favorite videos.
How It Works - Prerequisites - Installation Guide - Future Development - Contributing - License
Ziro is an all-in-one subtitle automation tool powered by advanced AI. It seamlessly handles the entire workflow - from accurately transcribing speech in your favorite videos and translating it into any language, to synchronizing the text and burning the final subtitles directly onto your video - delivering a ready-to-share, subtitled file in just a few clicks.
Our automated pipeline:
- Transcribes speech from your video using OpenAI's Whisper.
- Translates the text into Persian using Google Translate and DeepL.
- Generates & syncs perfectly timed subtitles.
- Renders the final video with subtitles burned in or as a separate file.
Before you begin, make sure you have the following installed:
- Python 3.11 or higher (for development)
- FFmpeg 8.0
- 8GB RAM (16GB recommended)
- GPU with CUDA support recommended for faster performance
Getting started with Ziro is quick and straightforward. Follow the steps below based on your operating system.
-
Download the Installer
Download the latest version of Ziro (.exefile) from the latest release page. -
Run as Administrator
Right-click the downloaded file and select “Run as administrator”. -
Install for All Users
When prompted, choose “Install for all users” and proceed by clicking Next until the installation is complete. -
Launch Ziro
After installation, Ziro will launch automatically.-
If you already have FFmpeg installed, the app will start loading its modules.
-
If FFmpeg is not installed, Simply run the following command in Command Prompt or PowerShell:
winget install "FFmpeg (Essentials Build)"Then restart Ziro.
-
-
You’re All Set!
Once the modules are loaded, you can start translating your favorite videos into any language subtitles in just a few clicks.
Support for Linux and macOS is coming soon.
Ziro currently operates in online mode only. An active internet connection is required for processing. Interruptions may affect the workflow.
Before processing, please choose your preferred Transcription and Translation models carefully.
On your first run, the selected models will be downloaded automatically in the background when you start the processing. Their approximate sizes are:
You won’t see a separate download progress bar – instead, you’ll notice the processing time is longer for the first run as the models are being fetched and installed.
| Transcription Model | Size (approx.) |
|---|---|
| Tiny | 70 MB |
| Base | default |
| Small | 460 MB |
| Medium | 1.4 GB |
| Large | 2.8 GB |
| Translation Model | Size (approx.) |
|---|---|
| Google Translate | online |
| DeepL | online |
Once processing finishes, you’ll see a success message. Click OK to automatically open the folder containing your subtitled video.
- Upgrade to PyQt 6 and UI improvements
- Add support for offline translation
Contributions are welcome! Please read our Contributing Guidelines to learn how you can help improve Ziro.ai.
This project is licensed under the AGPL-3.0 License - see the LICENSE file for details.
Third-Party Licenses:
- OpenAI Whisper: Licensed under the MIT License
- Google Translate: Subject to Google's Terms of Service
- DeepL: Subject to DeepL's Terms and Conditions