Phonon-2 for MLX

English speech recognition with Fermion Research's Phonon-2, derived from NVIDIA Parakeet TDT 0.6B v3. Preconverted five-value weights; approximately 350 MB. No converter or Fermion runtime needed. Requires mlx-audio with Phonon-2 support.

from mlx_audio.stt import load

model = load("mlx-community/Phonon-2")
result = model.generate("audio.wav", chunk_duration=30, overlap_duration=2)
print(result.text)

Supports token timestamps and overlapping chunk output, not live-input streaming. Validated against the official packed runtime on six clips using MLX 0.32.4.

Weights: CC-BY-4.0. Copyright Fermion Research and NVIDIA. Encoder codes are repacked into two MLX 2-bit planes; row palettes remain fp16 and dense tables use bf16. See NOTICE for attribution.