Runway's novel multimodal AI system can generate videos from text, images, and video clips.
Runway, the development team behind an eponymous Web-based machine-learning-powered video editor, has launched Gen-2, a brand-new multimodal AI system capable of generating videos using text, images, and video clips and a direct update to Gen-1, unveiled back in February.
According to the team, the novel system enables its user to realistically and consistently synthesize new videos, either by applying the composition and style of an image or text prompt to the structure of a source video, a feature that was present in Gen-1, or by using nothing but text prompts.
Gen-2 comes with three brand-new modes:
- Mode 1 – Text to Video: Synthesize videos in any style you can imagine using nothing but a text prompt.
- Mode 2 – Text + Image to Video: Generate a video using a driving image and a text prompt.
- Mode 3 – Image to Video: Generate video using just a driving image.
The AI also includes the modes introduced in Gen-1:
- Mode 4 – Stylization: Transfer the style of any image or prompt to every frame of your video.
- Mode 5 – Storyboard: Turn mockups into fully stylized and animated renders.
- Mode 6 – Mask: Isolate subjects in your video and modify them with simple text prompts.
- Mode 7 – Render: Turn untextured renders into realistic outputs by applying an input image or prompt.
- Mode 8 – Customization: Unleash the full power of Gen-1 by customizing the model for even higher fidelity results.