Modelpublic

theoracleguy/VibeVoice-Realtime-0.5B-fp16

sourceHugging Facemitupdated 26d agoView on Hugging Face

0likes69downloads

Model Card

VibeVoice-Realtime-0.5B-fp16

This is an MLX FP16 version of `microsoft/VibeVoice-Realtime-0.5B`, prepared for use with [OpenVox](https://openvoxai.com/) and other applications built on `mlx-audio`.

The model was converted to MLX format using mlx-audio 0.2.6.

For architecture details, supported features, licensing information, and other model-specific information, refer to the original Microsoft model card.

OpenVox

This model is provided for use with [OpenVox](https://openvoxai.com/), a local AI voice application for speech generation, voice cloning, audiobooks, transcription, dubbing, and other audio workflows.

OpenVox runs supported AI models locally on your device, allowing speech generation without requiring text or generated audio to be sent to a cloud speech service.

Website: https://openvoxai.com/

Use with mlx-audio

Install or update mlx-audio:

bash

pip install -U mlx-audio

CLI Example

bash

python -m mlx_audio.tts.generate \
  --model theoracleguy/VibeVoice-Realtime-0.5B-fp16 \
  --text "Hello, this is VibeVoice real-time 0.5B model." \
  --voice en-Emma_woman

Credits

—Original model: `microsoft/VibeVoice-Realtime-0.5B`
—Base model: `Qwen/Qwen2.5-0.5B`
—MLX inference framework: `mlx-audio`
—Packaged for use with [OpenVox](https://openvoxai.com/)