S

speech-synthesis

This is a fully client-side, JavaScript-only Text-to-Speech (TTS) service using eSpeakNG. No server-side bullshit. Just clone this repo into your desired directory, and you're all set.

Qwen3-TTS over SSH. Pick a voice, clone a voice, design a voice - all through a YAML config piped via stdin. Models run locally, no API keys, no cloud bullshit.

Give your AI agent a voice. Local speech stack for Apple Silicon: TTS, ASR, forced alignment, voices, daemon, and MCP bridge.