ποΈ Audio8-TTS: Free & Fast Zero-Shot Voice Cloning in Google Colab
This repository contains an easy-to-use Google Colab notebook for running Audio8-TTS. Generate realistic speech and perform zero-shot voice cloning using a short reference audio clip.
Runs in Google Colab with a compatible GPU.
π₯ Watch the Tutorial: Fastest Voice Clone for FREE | Text to Speech AI by @CoinNoin
π Run in Colab: Open Google Colab Notebook
β¨ Features
Multi-Language Support
- Cantonese
- Chinese
- Dutch
- English
- French
- German
- Italian
- Japanese
- Korean
- Polish
- Spanish
Two Model Options
Audio8-TTS-Preview-0.6Bβ High QualityAudio8-TTS-Preview-0.1Bβ Fast/Hybrid
Speech Generation
- Standard Text-to-Speech
- Zero-Shot Voice Cloning
- Reference audio-based voice generation
- Multilingual speech synthesis
π οΈ How to Use
- Open the Google Colab notebook.
- Go to Runtime β Change runtime type.
- Select T4 GPU if available.
- Run the dependency installation cell.
- Select your preferred model (0.6B or 0.1B).
- Run the model loading cell.
- Enter your target text to generate speech.
- For voice cloning, provide the exact transcript of your reference audio.
- Upload your reference
.wavfile. - Generate and download your audio.
π€ Credits
- Tutorial & Notebook Creator: CoinNoin
- Underlying AI Model: Audio8-TTS
β οΈ Responsible Use
Only clone voices that you own or have explicit permission to use. Do not use generated voices for impersonation, fraud, or misleading content.
Check the underlying model's license and usage restrictions before commercial use.