Grok Video 1.5
xAI · AI videoGrok Imagine Video 1.5 is xAI's preview image-to-video model. It animates a still image into a short clip with realistic motion, object interactions, and automatically generated synchronized audio—background music, sound effects, and ambience that match the scene.
Unlike Grok Imagine Video 1.0, this preview is image-to-video only: every generation needs a starting frame. Duration is 1–15 seconds at 480p or 720p, with common aspect ratios including 9:16 and 16:9.
On VidMachine, scene generation uses each scene's starting frame and motion prompt at 720p, billed at 16 credits per second.
Key features and benefits
Image-to-video with native audio
Upload a portrait, product still, or illustration and describe the motion. The model generates picture and sound in one pass, so clips include matching ambience, effects, and optional short dialogue without a separate audio step.
Cinematic motion
Prompts work best when they direct action and camera rather than re-describing the image. The model understands pan, tilt, dolly, orbit, handheld, and intensity modifiers so you can specify how the subject and camera should move.
Flexible clip length
Generate 1–15 second clips. Shorter 5–8 second runs tend to be more stable; longer clips are available when a scene needs more time.
Technical specifications
ProviderxAI (Grok Imagine Video 1.5 preview)
ModeImage-to-video only (starting frame required)
Resolution480p or 720p (VidMachine uses 720p)
Duration1–15 seconds
Aspect ratiosauto, 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, 2:3
AudioNative synchronized audio
Use cases and applications
Use Grok Video 1.5 when you already have a strong key frame and want it animated with matching sound—product showcases, portrait motion, character stills, and 9:16 social clips.
It is a natural upgrade from Grok Video 1.0 when native audio and the 1.5 motion profile matter more than the lower 1.0 credit rate.
Why this model
Pick 1.5 when you want xAI's latest Imagine video preview with built-in audio. Stay on 1.0 when you need text-to-video without a start frame or a lower per-second cost.
Compare look and credit cost with other 720p models in project settings before a long export.
How VidMachine uses it
VidMachine generates scene clips via image-to-video at 720p, using each scene's starting frame and motion prompt. Duration is clamped to 1–15 seconds. Generation is routed through Replicate.
Billing is 16 credits per second of output video.
What you should know
How is 1.5 different from Grok Video 1.0?
1.5 is a preview image-to-video model with native synchronized audio. 1.0 supports text-to-video as well as image-to-video and is billed at a lower credit rate. On VidMachine you can switch between 1.0 and 1.5 in the video model selector.
Does every clip need a starting image?
Yes. Grok Imagine Video 1.5 requires an input image. VidMachine always sends the scene starting frame.
What resolution does VidMachine use?
Scene generation runs at 720p, billed at 16 credits per second.