Amphion
The most comprehensive audio AI research toolkit, offering TTS, singing voices, music, and sound effects in one place
📋 Info
| GitHub Stars | ⭐ 9.8k Stars |
| License | MIT |
| Language | Python |
| Version | 0.1.1-alpha |
| Updated | 2026-05-10 |
📖 Overview
Amphion is the most comprehensive open-source audio AI research toolkit developed by OpenMMLab (9.8k Stars). It offers a one-stop solution for all types of audio generation tasks, including text-to-speech (TTS), singing text-to-speech (SVS), music creation, sound effect generation, and voice conversion. Designed for researchers, it features a clean code structure that ensures experiment reproducibility. The toolkit supports over 30 different model architectures, making it suitable for audio AI researchers and projects that require diverse audio generation capabilities.
✨ Features
- One-click TTS combined with singing voices, music, and sound effects in one solution.
- Support for 30+ model architectures
- Designed for research purposes — reproducible
- Rich evaluation metrics + visual / visualization
- OpenMMLab quality assurance
Advertisement