Grok Imagine Video 1.5

xAI announced Grok Imagine Video 1.5 in preview on June 3, 2026, naming the model key grok-imagine-video-1.5-preview and describing it as an image-to-video model now available via the xAI API. The model takes a single still image plus a natural-language prompt and animates the scene, preserving the source image's detai According to the xAI post, users can direct the shot by describing the camera move, pacing, and sound design in plain language, then setting resolution and clip length. The announcement includes a short Python code snippet demonstrating how to animate an image in a few lines via the API. This is a first-party xAI relea

Vercel's AI Gateway playground for Grok Imagine Video 1.5 Preview exposes the model's parameter surface for image-to-video generation. Pricing is set at $0.08 per second at the lowest available configuration, with billing tied to seconds of generated output. Developers must supply a required start frame image and may a The playground exposes duration from 1s to 15s, resolution at 480p, 720p, and 1080p with an Auto option, aspect ratios including 1:1, 16:9, 9:16, 4:3, 3:4, 2:3, and 3:2, and batch generation of 1 to 4 videos per request. Access to the model is gated behind an API key obtained from Vercel, with use subject to SpaceXAI's

Vercel AI Gateway lists Grok Imagine Video 1.5 Preview, the preview build of SpaceXAI's Imagine Video 1.5 image-to-video generation, released on May 30, 2026. The model animates a starting image with synchronized audio in a single pass and is documented as improving audio quality, prompt following, and photorealism ove Integration runs through the experimental generateVideo function from AI SDK 6 or later, where developers pass a prompt object with an input image plus a text motion description. Parameters expose duration, aspect ratio, and resolution at 480p, 720p, and 1080p, with billing per second of generated video (rates by resol