Gen-3 Alpha is a text-to-video and image-to-video model from Runway, announced on June 17, 2024. Runway had been an early mover in commercial AI video with its Gen-1 and Gen-2 systems, and Gen-3 Alpha was its next-generation foundation model, trained on new infrastructure built for large-scale multimodal training on both video and images.
Runway described Gen-3 Alpha as a major improvement in fidelity, consistency, and motion over Gen-2, and notably framed it as a step toward building general world models. The model was trained on highly descriptive, temporally dense captions, which the company said enabled fine-grained control over timing and key-framing, and it was reported to generate expressive human characters with a wide range of actions and emotions. It powered tools including text-to-video, image-to-video, Motion Brush, advanced camera controls, and a Director Mode, and shipped with an in-house moderation system and C2PA content-provenance support.
Gen-3 Alpha was significant both as a competitive product and for the language Runway used around it: even a commercial creative tool was being positioned as progress toward models that understand how the world works. For a general reader, it captures how the video-generation race in 2024 was reframed from making nice clips to building simulators of reality, while the practical product remained a creative tool for filmmakers and marketers.