Text-to-speech tool that synthesizes natural speech from short voice samples.
What is Fish Audio?
Fish Speech is a text-to-speech (TTS) tool developed by the creator of So-VITS-SVC and Bert-VITS2. It can synthesize natural, fluent speech from 15 seconds of audio of any voice while preserving the original voice’s timbre, style, and accent. Fish Audio is an audio generation platform that provides users with a variety of voice models to discover and use.
How do I use Fish Audio?
Users can discover and use pre-built speech models or create their own. The platform provides a text-to-speech toolkit that allows users to enter text and select a speech model to generate speech.
Fish Audio's core features
Text-to-speech synthesis
Speech Model Discovery
Building Custom Speech Models
Preserve the timbre, style, and accent of the original voice.