whisperkit-cli

On-device speech-to-text (Whisper) + text-to-speech (Qwen3-TTS) CLI. Runs on the Apple Neural Engine (ANE), Apple's low power, dedicated ML inference chip. Models download once on first run, then all inference is local.