Real-Time Video Generation Researcher

Cantina Labs is seeking a researcher to develop real-time video generation models for interactive characters, focusing on realistic humans, stylized figures, and non-human identities. The role involves advancing character expression, motion coherence, and speech-to-movement relationships while optimizing for latency, quality, and efficiency. You will work across streaming generation, fine-tuning, distillation, and data development to create responsive and consistent AI-driven characters.