Instructions to use Qwen/Qwen3-ASR-0.6B-hf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Qwen/Qwen3-ASR-0.6B-hf with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="Qwen/Qwen3-ASR-0.6B-hf")# pip install -U transformers accelerate # Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("Qwen/Qwen3-ASR-0.6B-hf") model = AutoModelForMultimodalLM.from_pretrained("Qwen/Qwen3-ASR-0.6B-hf", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Off topic: Qwen 3.8 license suggestion
I am sorry for reaching out this way, but I'm not really sure of a better way honestly. There are rumors Qwen 3.8 will be licensed in such a way that will require providers making money from Qwen to give a share of profits. I think this is completely reasonable personally and is great for supporting/incentivizing future open model development, but it has caused a lot of controversy, with people claiming Qwen is moving away from open source (not true in my opinion).
I just wanted to suggest that maybe there could be a clause in the license that this restriction doesn't apply or becomes a less restrictive license like MIT after a certain date. I am fairly certain this is legally doable and would likely build trust while achieving the same thing. Please consider. Thanks, best of wishes for the release :-)