Moshi is a speech-text foundation model and full-duplex spoken dialogue framework.
Run and fine-tune open-source models. Deploy custom models at scale. All with one line of code.
The platform where the machine learning community collaborates on models, datasets, and applications.
