It provides enhanced stability and emotional expression capabilities, and can clone anyone's voice with just a 10-second audio prompt!

Let's take a look at the results first:

Input Sample（Nahida | Genshin Impact）：

Synthesized voice:

The lights of the human world are reflected in the lake, her longing stirs ripples in the still water. If the price is only loneliness, then let this wish flow freely. Flow into the world she gazes upon, and also into her gaze as clear as lake water.

has the following features:

• Trained with 7 million hours of multi-language data (a significant increase from the previous 200,000 hours)
• Now supports 8 languages: English, Chinese, German, Japanese, French, Spanish, Korean, and Arabic
• Fully open-source, providing support for developers and researchers worldwide

Main functions:

• Ultra-low latency high-speed TTS (Text-to-Speech)
• Instant voice cloning
• Supports local deployment or cloud services

Everyone can try it out on the official website: https://fish.audio