I'm not connected to the makers of OpenVoice (MyShell), I'm just a solo developer and founder of Valyrian Tech.

I'm building my own chatbot that has multiple personas and I wanted each persona to have a unique sounding voice. I needed a simple way to handle the text-to-speech feature, to be more specific, I needed a way to make a simple GET request and get an audio file spoken with a specific sounding voice, all in less than a second to make fluent conversations possible.

OpenVoice is great for this, but since it is more a research project than a commercial product, there was no easy API available, at least not with the functionality I needed, so I made this simple API server.

What this is good for: Chatbots where you need a custom voice in multiple languages or accents in sub-second generation times.

What this is NOT good for: Cloning existing voices to fool people, it's just not going to sound convincing, there are more advanced solutions available.

There is also an easy one-click template available on Runpod for those who want to try it out.