Grok

Grok Imagine Video is xAI's video generation model, powered by the proprietary Aurora autoregressive engine. Initially launched as part of the Grok ecosystem on X (formerly Twitter), the model became publicly available via API and partner platforms in February 2026. It supports text-to-video, image-to-video, and video editing capabilities.

The model generates video at 720p or 480p resolution, at 24 frames per second, with durations from 1 to 15 seconds. Aspect ratio support is extensive: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, and 2:3 are all available, making Grok Imagine flexible across social media formats. Average generation time is approximately 30 seconds, which is among the fastest in the industry.

A standout capability of Grok Imagine Video is instruction following. The model excels at accurately interpreting complex text prompts with multiple subjects, actions, and scene elements. xAI reports that Grok Imagine has generated over 1.2 billion videos in a single 30-day period, indicating massive real-world adoption. The March 2026 "Extend from Frame" update added the ability to chain clips using the final frame of one generation as the start of the next, enabling sequences up to 15 seconds per clip.

Grok video generation on ZenCreator provides access to this model with Start Frame support for image-to-video workflows. The partial content filters make it more permissive than fully filtered alternatives, while xAI continues to develop the model toward longer-form content — with stated goals of 30-minute generation by late 2026.