xAI has launched a preview version of its image-to-video model, Grok-Imagine-Video-1.5-preview, through its API. The model enables users to convert a single still image into a fluid, cinematic video by providing a starting frame and a prompt that describes the motion. The output includes camera moves, atmosphere, and physics, while remaining faithful to the source image. Users can generate clips at up to 720p resolution, with the ability to direct the shot using natural-language prompts. The model maintains the detail and lighting from the input frame, ensuring continuity with the original image rather than reinterpreting it. It also supports sequences, allowing users to stage each frame, animate it, and chain the shots together into longer scenes that maintain a consistent visual style across an entire project. Source: xai
The model works by taking a starting frame and a prompt that describes the motion, then animating the scene with camera moves, atmosphere, and physics. It preserves the detail and lighting from the input frame, so the result continues the original image rather than reinterpreting it. The model can generate clips at up to 720p resolution, with users able to set the resolution and clip length. It also supports sequences, allowing users to stage each frame, animate it, and chain the shots together into longer scenes that maintain a consistent look across an entire project. Source: xai
The release comes as xAI continues to expand its capabilities in generative AI, offering developers access to the model via its API. Users can generate video clips using a few lines of code by importing the xai_sdk and using the video.generate method with the specified parameters. Source: xai
