Multimodal model

Synthesia Express-1

Express-1 is Synthesia’s closed-source avatar video model for generating expressive AI avatar performances with matched voice tone, body language, facial expression, and lip sync.

Type
Multimodal
Weights
Closed