xAI Launches Grok Imagine 1.5 Preview Image-to-Video Model

xAI has unveiled the Grok Imagine Video 1.5 Preview, the latest iteration of its image-to-video AI model, now accessible through its API. Announced on June 3, 2026, this release builds on Grok Imagine 1.0’s foundation, introducing enhanced video quality, extended workflows, and production-ready support for developers.

Key upgrades include the ability to transform a single image into fluid, cinematic video clips at up to 720p resolution. Users can direct animations using natural-language prompts, specifying camera movement, pacing, and even sound design. The model retains the input image’s details and lighting, ensuring continuity rather than reinterpretation. Developers can also chain multiple clips together, creating consistent, longer sequences—a feature particularly useful for content creators and filmmakers.

According to recent third-party evaluations, Grok Imagine 1.5 has outperformed competing models, such as Seedance 2.0 and HappyHorse 1.0, topping the Arena.ai Image-to-Video leaderboard with a +52 Elo point gain as of May 30, 2026. This improvement in visual quality and user-directed workflows positions Grok Imagine 1.5 as a leader in the image-to-video AI category.

The release is part of xAI’s broader Grok ecosystem, which has seen rapid evolution since its inception. Grok-1, launched in 2023, introduced a 314-billion-parameter Mixture-of-Experts model, while Grok Imagine 1.0 debuted earlier this year, offering 10-second video generation capabilities. The 1.5 update further refines xAI’s offerings, integrating features like “Video Extend” for chaining animations and a standardized API for seamless developer access.

Beyond its technical capabilities, Grok Imagine 1.5 also reflects xAI’s ambitions to dominate both commercial and governmental AI markets. Notably, xAI secured a deal with the U.S. Department of Defense in February 2026 to implement Grok models in classified systems, highlighting the strategic importance of the technology.

Developers interested in experimenting with the model can get started with a few lines of code via the API, as demonstrated in the preview announcement. The model’s ability to generate cinematic animations with minimal input makes it a compelling tool for industries ranging from media production to virtual reality.

As of now, Grok Imagine Video 1.5 Preview is available in its initial API rollout, with further updates expected as xAI continues to refine its platform. For developers and enterprises looking to leverage cutting-edge AI-driven video generation, this release could mark a significant step forward in automation and content creation.

Image source: Shutterstock