kyutai-labs/moshi
Daily ranking profile for kyutai-labs/moshi in AI Rank.
Moshi is a speech-text foundation model and full-duplex spoken dialogue framework. It uses Mimi, a state-of-the-art streaming neural audio codec.
- Stars: 10202
- Forks: 952
- Language: Python
- Category: Models
- Subcategory: Audio/Speech models
- Keywords: Speech-to-text, Text-to-speech