#19

Runway

Runway develops generative video and world models, including the Gen-4/Gen-4.5 text-to-video series and GWM-1, a frame-by-frame predictive world model with variants for robotics simulation, interactive environments, and avatar synthesis. Research spans diffusion-based video generation, post-training for temporal consistency, and multimodal control across image, video, and audio modalities.