Custom Text-to-Speech integration using MiniMax API (speech-02-turbo and newer models) via WebSocket for minimal latency.
I previously used ElevenLabs for many months, but they don’t allow cloning of any voice (e.g., a famous person’s voice), so I looked for an alternative and MiniMax Audio met my requirements. There was no Home Assistant integration, so I wrote my own.
It’s a first public release so it can contain some bugs.