Grok Imagine 1.5 animates images with AI via API
xAI has put its `grok-imagine-video-1.5-preview` model into preview, allowing you to turn a still image into a video of up to 720p using natural-language instructions. You can also animate multiple frames and chain them together to create longer scenes with a consistent style.
xAI now lets you turn a still image into a video clip with grok-imagine-video-1.5-preview, its new image-to-video model available through the xAI API as a preview.
You only need to provide an opening frame and describe in words what should happen: camera movement, the scene's pacing, its atmosphere or even sound design cues. The model generates a video of up to 720p and aims to preserve the original image's details, lighting and style.
From a still image to a moving scene
The idea is simple. If you have an illustration of a city in the rain, you can ask the camera to move slowly down the street, have the lights reflect on the pavement and make the atmosphere more cinematic. Grok Imagine 1.5 animates the scene without completely reinterpreting the source image.
That matters for work where visual consistency is important. Instead of generating every frame from scratch, the model uses the initial image as a reference to preserve its appearance while adding movement, camera changes and environmental elements.
You can define the duration and resolution through the API, along with the shot instructions. This makes it possible to automate clip creation from image catalogs, storyboards or previously prepared designs.
It can also be used to join multiple shots
The model is not limited to a single clip. You can prepare several frames, animate them separately and chain the shots together to create longer scenes with a consistent look.
That workflow can be useful for:
- Turning illustrations into video pieces.
- Creating quick animations for social media.
- Testing camera movements before production.
- Bringing storyboards and visual concepts to life.
Availability is still a preview, so it should not be confused with a final version or a guarantee of identical results in every case. Access is also offered through the xAI API, a format aimed mainly at developers and teams that want to integrate video generation into their own tools.
For you, the practical change is that a single image can become the starting point for a complete sequence, without having to animate every element manually. The question now is how well it maintains that consistency as clips get longer and scenes are chained together.