Mixer AI
Grok Imagine Video
Animate a frame, try several ideas and build a clip lasting 6–30 seconds.
When to choose it
Use Grok Imagine Video for early experiments: see whether a movement, angle or concept works. Start with a photo or a description and give it one main action. For longer clips, set out consecutive beats so the character does not restart the same action halfway through.
- Short experiments and clips up to 30 seconds share one model page.
- Multiple images can guide a scene when the selected resolution supports them.
Modes and settings
- Text or photos, 6–30 seconds, at 480p, 720p or 1080p.
- Up to 7 photos at 480p and 720p; one at 1080p. Presentation modes include Normal, Fun and Spicy.
Price in coins
| Quality / size | Input | Coins / second |
|---|---|---|
| 480p | First frame | ≈ 1.65 |
| 480p | Text | ≈ 1.65 |
| 720p | First frame | ≈ 2.99 |
| 720p | Text | ≈ 2.99 |
| 1080p | First frame | ≈ 5.12 |
| 1080p | Text | ≈ 5.12 |
Billing is per second, not per fixed-length clip. The rate is multiplied by the billable duration; the complete request total is rounded up to a whole coin once.
≈ marks a fractional rate rounded to two decimal places for display; billing uses full precision.
What you can make
- A vertical social clip from a portrait.
- Trying camera moves before a more expensive final generation.
How to direct the result
- Name the action, camera direction and final state.
- Explain the role of each photo instead of expecting an automatic slideshow.
Example: a portrait in the breeze
6 seconds. Animate the portrait: a light breeze moves the hair and collar. The person looks calmly into the distance, then toward the camera. A gentle push-in. Preserve facial features, clothing and background. Restrained motion, no scene change.
What to keep in mind
- Presentation modes do not override service restrictions or guarantee particular content.
- Inspect identity, clothing and props throughout longer clips.
Sources
- Prices and availability — Mixer AI catalog, updated 2026-09-18.
- Official maker site: xAI
Model facts verified: 2026-09-10.
Guides and comparisons using this model
FAQ
Why does 1080p allow only one photo?
That is the input limit for this mode. Choose 480p or 720p to use several images.
Can I start from text alone?
Yes. Describe the entire scene, not just movement: without a photo, appearance is not defined.
How long should a first test be?
Six seconds is enough to check one action without paying for a long unsuccessful scene.
Do Normal, Fun and Spicy set the resolution?
No. They are presentation modes; resolution is selected separately.
Can I extend a successful clip?
Eligible results support extension. It is a separate request with its own price and settings.
Why does a thirty-second scene repeat itself?
Give it consecutive beats and an ending. Avoid open-ended instructions such as walking indefinitely.
What changes the price?
The coin rate per second depends on quality, mode and, for some models, sound. Multiply it by the billable duration. Where input video is charged, its duration also counts. The complete request total is rounded once.
Do I need a separate model subscription?
No. In Mixer AI, you top up one coin balance and use it for the tasks you need. You do not need a separate subscription to this model.