Gen-2 by Runway

Gen-2 is a multimodal AI system developed by Runway Research that generates novel videos through multiple creative modes. Users can synthesize videos using text prompts alone (Text to Video), combine driving images with text prompts (Text + Image to Video), or transform existing images and video clips into new video content. The system offers seven distinct generation modes including stylization, storyboard conversion, rendering, and customization options for enhanced fidelity. Gen-2 is designed for filmmakers, creative professionals, and content creators who want to produce video content without traditional filming equipment. The tool enables rapid video prototyping and exploration of creative ideas by converting conceptual descriptions into visual footage. According to user studies, Gen-2 outputs were preferred over Stable Diffusion and Text2Live for image-to-video and video-to-video translation tasks. One notable limitation is that while the tool offers various generation modes, the quality and realism of outputs may vary depending on the specificity and clarity of input prompts or source materials.