Grok Imagine

What it does

Grok Imagine is xAI's image-to-image model: you supply an existing image plus a plain-language instruction, and it returns an edited version of that image. It does not create pictures from a blank prompt. A source image is always the starting point. On Pipe2 it runs in the Image Editor pipeline, and edits can be produced square (1:1), landscape (16:9), portrait (9:16), or in 4:3 and 3:4.

When to use it

• Run it when you already have a picture and want to change its style, restyling a photo without redrawing it from scratch.
• Use it to remove an object from a shot, or to modify one element while leaving the rest of the frame intact.
• Pick it when the whole look of an image needs transforming and you would rather describe the change in a sentence than mask and layer it by hand.
• Choose an aspect ratio to match the destination: 9:16 for vertical feeds, 16:9 for landscape, 1:1 for square placements.
• Reach for Grok Imagine Quality instead when you need higher fidelity and the extra cost per edit is worth it.

What to know before you run it

Grok Imagine is metered as a single unit (a Grok Imagine image edit), so there are no resolution or duration tiers to choose between. What drives your total is how many edits you run, so plan for iteration when a prompt needs several passes. The model is image-to-image only, meaning a source image is mandatory; there is no text-to-image path here. If you want to edit a video rather than a still, that is a different model: Grok Imagine Video, which runs in the Video Editor pipeline.

Also known as

You may see this model referred to as Grok Imagine 1.0 or xAI Imagine. Its identifier on Pipe2 is grok-imagine-image, matching xAI's own model id, and the metered unit on your billing line reads Grok Imagine image edit.