Download Latest Version
v0.4.2 source code.tar.gz (4.5 MB)
Get an email when there's a new version of MLX-Audio
| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| README.md | 2026-03-30 | 3.7 kB | |
| v0.4.2 source code.tar.gz | 2026-03-30 | 4.5 MB | |
| v0.4.2 source code.zip | 2026-03-30 | 4.8 MB | |
| Totals: 3 Items | 9.3 MB | 2 | |
What's Changed
- fix: pin setuptools<81 to prevent webrtcvad import failure by @mm65x in https://github.com/Blaizzy/mlx-audio/pull/579
- docs(whisper): add distil-whisper to supported models list by @mm65x in https://github.com/Blaizzy/mlx-audio/pull/580
- Fix whisper timestamps by @Blaizzy in https://github.com/Blaizzy/mlx-audio/pull/583
- add moshi sts by @mm65x in https://github.com/Blaizzy/mlx-audio/pull/569
- Fix Qwen3-TTS streaming memory leak. by @orbitalquark in https://github.com/Blaizzy/mlx-audio/pull/585
- Add Trendshift badge to README by @Blaizzy in https://github.com/Blaizzy/mlx-audio/pull/588
- Fix Parakeet streaming defaults and add validation by @andimarafioti in https://github.com/Blaizzy/mlx-audio/pull/485
- fix(fish_qwen3_omni): support loading quantized/converted models in sanitize() by @yoshphys in https://github.com/Blaizzy/mlx-audio/pull/584
- fix(parakeet): align preprocessing with NeMo reference by @beshkenadze in https://github.com/Blaizzy/mlx-audio/pull/592
- fix: convert numpy audio to mx.array in realtime STT streaming path by @korale77 in https://github.com/Blaizzy/mlx-audio/pull/596
- Add Irodori-TTS: Japanese TTS model port to MLX by @yoshphys in https://github.com/Blaizzy/mlx-audio/pull/591
- fix(vibevoice): preserve quantization metadata in sanitize() for quantized model loading by @korale77 in https://github.com/Blaizzy/mlx-audio/pull/604
- fix: handle float voice parameter in KokoroPipeline (fixes [#224]) by @sudabg in https://github.com/Blaizzy/mlx-audio/pull/603
- Add Cohere Transcribe ASR model by @lucasnewman in https://github.com/Blaizzy/mlx-audio/pull/605
- Add Voxtral-4B-TTS-2603 support by @shreyaskarnik in https://github.com/Blaizzy/mlx-audio/pull/607
- Restore *.pth in download patterns for VoxCPM by @shreyaskarnik in https://github.com/Blaizzy/mlx-audio/pull/609
- Add HumeAI Tada by @Blaizzy in https://github.com/Blaizzy/mlx-audio/pull/610
- docs: move Voxtral TTS from STT table to TTS table by @hewliyang in https://github.com/Blaizzy/mlx-audio/pull/615
- fix(tts): allow streamed generation to save audio to disk by @mvdirty in https://github.com/Blaizzy/mlx-audio/pull/608
- Add KugelAudio TTS model by @n0madic in https://github.com/Blaizzy/mlx-audio/pull/614
- Add Qwen2-Audio-7B-Instruct support by @shreyaskarnik in https://github.com/Blaizzy/mlx-audio/pull/617
- Move kugel tests by @Blaizzy in https://github.com/Blaizzy/mlx-audio/pull/622
- fix(stt): remove PyTorch dependency from Cohere ASR buffer loading by @isshintian-cohere in https://github.com/Blaizzy/mlx-audio/pull/616
- feat(voxtral_tts): add streaming support with overlap-add decoding by @chazmaniandinkle in https://github.com/Blaizzy/mlx-audio/pull/618
- Bump version to 0.4.2 by @Blaizzy in https://github.com/Blaizzy/mlx-audio/pull/625
New Contributors
- @yoshphys made their first contribution in https://github.com/Blaizzy/mlx-audio/pull/584
- @korale77 made their first contribution in https://github.com/Blaizzy/mlx-audio/pull/596
- @sudabg made their first contribution in https://github.com/Blaizzy/mlx-audio/pull/603
- @hewliyang made their first contribution in https://github.com/Blaizzy/mlx-audio/pull/615
- @mvdirty made their first contribution in https://github.com/Blaizzy/mlx-audio/pull/608
- @n0madic made their first contribution in https://github.com/Blaizzy/mlx-audio/pull/614
- @isshintian-cohere made their first contribution in https://github.com/Blaizzy/mlx-audio/pull/616
- @chazmaniandinkle made their first contribution in https://github.com/Blaizzy/mlx-audio/pull/618
Full Changelog: https://github.com/Blaizzy/mlx-audio/compare/v0.4.1...v0.4.2
