Forty-eight percent of test subjects mistook Tavus's new Griffin model for a real human during a one-minute video call. Previous systems only achieved a two percent success rate. The model synchronizes facial expressions, tone, and gestures in real time.