Modelpublic

mlx-community/Qwen3-ASR-0.6B-8bit

sourceHugging Faceapache-2.0updated 8mo agoView on Hugging Face

6likes220kdownloads

Model Card

mlx-community/Qwen3-ASR-0.6B-8bit

This model was converted to MLX format from `Qwen/Qwen3-ASR-0.6B` using mlx-audio version 0.3.1.

Refer to the original model card for more details on the model.

Use with mlx-audio

bash

pip install -U mlx-audio

CLI Example:

bash

python -m mlx_audio.stt.generate --model mlx-community/Qwen3-ASR-0.6B-8bit --audio "audio.wav"

Python Example:

python

from mlx_audio.stt.utils import load_model
from mlx_audio.stt.generate import generate_transcription

model = load_model("mlx-community/Qwen3-ASR-0.6B-8bit")
transcription = generate_transcription(
    model=model,
    audio_path="path_to_audio.wav",
    output_path="path_to_output.txt",
    format="txt",
    verbose=True,
)
print(transcription.text)