Docker image to run a self-hosted Kokoro TTS server with an OpenAI-compatible audio speech API. Supports 20+ voices and NVIDIA GPU (CUDA) acceleration. Simple drop-in replacement for OpenAI TTS API. - View it on GitHub