Post

Running speech and music models locally on a Mac usually means slow CPU inference, but mlx-audio runs TTS, STT, STS and music generation fast on Apple Silicon via MLX.