What is Grok Imagine? Image and video generation, plans and how to use it

Grok Imagine is the part of Grok that makes pictures and video. You type what you want to see, or start from a photo, and it generates images, animates them into video with sound, and edits what it made. In the past three months SpaceXAI has shipped three big updates: Video 1.5, references for video, and Image 2.0. This guide explains what it can do, which Grok plans include it, how to start and what to watch out for.

Grok Imagine at a glance

What’s new: three updates in 2026

What Grok Imagine can do

Images. Image 2.0 is built for images you can use in real work. SpaceXAI says it follows instructions closely, plans typography and layout, keeps small text sharp and preserves what you put in across generations and edits. Its editing tools include:

• Magic wand: edits the region you point at and leaves the rest untouched.
• Segmentation: selects precise areas to change.
• Background removal: exports a subject with a transparent background.
• Multi-reference editing: accepts up to five input images in one generation, so you don’t have to composite by hand.
• Smart resize: pick a ratio and the model fills in the frame.
• Templates: ready-made starting points for photo edits, product shots, headshots, icons, game assets and merchandise.

Video. Video 1.5 turns a prompt or a starting image into a clip with generated sound: effects, ambience and dialogue. With the July 31 update you can describe a shot with no starting image, ask for native 1080p, and lock characters, scenes or products in place with up to seven reference images. Voice references keep the same voice from scene to scene.

Workflow. Projects organise your work in the sidebar, you can run several prompts at once with multiple agents, and you can search your library of images and videos.

Which Grok plans include Imagine?

SpaceXAI’s pricing page shows:

Image and voice references started with SuperGrok Heavy and SuperGrok Plus in the US on July 31, before rolling out to all tiers. For everything else that comes with Grok, see the Grok profile.

How to use Grok Imagine

1. Open it. Go to grok.com/imagine, or open the Grok app on iOS or Android and choose Imagine.
2. Choose image or video. Describe what you want, or upload a starting image.
3. Pick the quality. For images, Quality Mode uses Image 2.0. For video, SpaceXAI’s Video 1.5 Fast makes a 6-second, 720p clip in about 25 seconds.
4. Refine. Use the magic wand or segmentation to change one area, remove the background, or resize to a new ratio.
5. Add references. For video, add reference images and, if you have access, a voice reference to keep a character consistent.
6. Organise. Save work in a Project and use search to find it later.

Menu names change often, so follow the labels in the app.

Grok Imagine vs ChatGPT Images

Read our guide to ChatGPT Images 2.5 and our Grok vs ChatGPT comparison, or see how it stacks up against Midjourney and Adobe Firefly in our roundup of the best AI image generators.

Safety and responsible use

Grok’s image tools have been misused. In January 2026, CNBC reported that xAI faced backlash after Grok generated sexualised images of children on X, and later that xAI limited Grok’s ability to create sexualised images of real people on X. Don’t create or share images of real people without their consent, and check SpaceXAI’s policies before using generated images commercially.