SeamlessM4T

SeamlessM4T is a foundational AI model developed by Meta FAIR for multilingual speech translation. It supports translation of nearly 100 input languages and 35 output languages. The model enables speech-to-speech and speech-to-text translation while preserving expressive elements. A related model, SeamlessExpressive, focuses on maintaining speech style and emotion in translations. The demo allows users to try both models for input in supported languages.

Key Benefits

• Supports nearly 100 input languages
• Supports 35 output languages
• Foundational model for speech translation