30-Second Brief
The News: Grok has launched a multi-image to video feature, letting users combine up to seven images into a single generated video clip with improved control over characters and settings.
Why It Matters: X Premium subscribers now have a significantly more powerful AI video creation tool built directly into the apps they already use β no third-party service required.
Source: @grok on X
Grok Multi-Image to Video Is Live: Here's How to Use It
Grok just shipped one of its most capable creative features yet. As of March 13, 2026, the multi-image to video tool is live across iOS, Android, and the web β and it's a meaningful step up from single-image generation. Instead of feeding Grok one photo and hoping for the best, you can now chain up to seven separate images into a single cohesive video, with tighter control over character consistency and scene settings throughout.
π What Changed
| Feature | Before | Now |
|---|---|---|
| Image inputs | Single image | Up to 7 images |
| Character consistency | Limited | Improved control across frames |
| Output resolution | 720p | 720p with synchronized audio |
| Max clip length | Standard segment | Up to 15 sec via Extend from Frame |
| Platform availability | Web only | iOS, Android, and Web |
| Access requirement | X Premium | X Premium (unchanged) |
How the Feature Actually Works
The multi-image to video tool is powered by xAI's Aurora engine, running on 110,000 NVIDIA GB200 GPUs. In practice, you're giving Grok a visual storyboard β up to seven images representing characters, locations, or objects β and the model weaves them into a single continuous video. The key improvement in this update is character and setting consistency: Grok now does a better job of keeping your subjects recognizable across frames rather than drifting into different-looking versions of the same person or scene.
If you want longer clips, the Extend from Frame feature (launched March 2, 2026) lets you chain segments together up to 15 seconds each. Output comes in 720p with audio baked in β no silent clips that need separate audio work.
For developers, the Grok Imagine API supports image-to-video capabilities at $0.05 per second for 720p video with audio, having launched on January 28, 2026.
π¦ Owner's Action Plan
Verdict: RECOMMENDED β Worth trying today if you have X Premium
- Confirm your X Premium subscription β This feature requires X Premium. If you're not subscribed, you won't see the option.
- Open Grok on your preferred platform β Available now on iOS, Android, and web. No update required if you're on the latest app version.
- Select the image-to-video tool β Look for the video generation option within Grok Imagine. The multi-image input will now accept up to seven images in a single session.
- Prepare your images β Choose up to 7 photos that share characters, locations, or objects you want to appear in the video. Consistency in your source images helps Grok maintain consistency in the output.
- Use Extend from Frame for longer clips β Once your initial clip generates, use the Extend from Frame option to chain additional segments if you need footage beyond the default length (up to 15 seconds per segment).
- For developers: test the API β If you're building on top of Grok Imagine, the image-to-video API endpoint is live at $0.05/second for 720p output with audio.
π° Deep Dive
The jump from single-image to multi-image input is more significant than it sounds. Single-image video generation has always had a ceiling: you're essentially asking the model to animate one moment. Multi-image input lets you define a narrative arc β a beginning, middle, and end β that Grok then interpolates into motion. The character consistency improvements are what make this practically useful rather than just technically interesting. Earlier AI video tools struggled badly with keeping faces and objects coherent across frames, which made them feel like novelties rather than real creative tools.
The combination of multi-image input, the Extend from Frame feature, and synchronized audio output puts Grok's video generation in a more competitive position among AI creative tools. The 720p ceiling is a real limitation for professional use cases, but for social content and personal projects, it's more than adequate. The fact that this runs inside the X/Grok app β rather than requiring a separate subscription to a standalone video AI platform β is a genuine convenience advantage for existing X Premium subscribers.
The API pricing at $0.05 per second is worth watching. At that rate, a 15-second clip costs $0.75 β reasonable for prototyping, and it gives developers a clear cost structure to build products around. As Aurora's capabilities expand and the model improves, that price point could make xAI's video generation a meaningful competitor to dedicated video AI APIs.
Sarah focuses on Tesla Energy, SpaceX missions, and the broader Musk AI portfolio. Former data analyst in clean energy. Based in San Francisco.
Sources verified at publish time. Spotted an inaccuracy? Email editorial@basenor.com.
