voxkit — Voice pipeline toolkit
Provides abstractions for building voice-powered apps in Rust:
- Audio types —
AudioChunk,Transcript,Segment - STT — Speech-to-text trait + OpenAI backend
- VAD — Voice activity detection trait + Silero backend
- TTS — Text-to-speech trait + OpenAI backend
- Mic — Cross-platform microphone capture via cpal
- System audio — macOS system audio capture via ScreenCaptureKit
- Mic mode — macOS microphone mode detection
- TTS player — Background TTS playback with sentence splitting
All backends are feature-gated. Core types have zero dependencies.