Users can now generate cinematic clips with native audio using a new lightweight video model available through third party providers. The Grok Imagine Video 1.5 Lite supports text-to-video and image-to-video generation for sequences between 1 and 15 seconds, producing output in 480p, 720p and 1080p resolutions across 7 different aspect ratios.

The model from SpaceXAI is currently deployed on the fal and Venice platforms. Venice offers users the ability to create content anonymously and states that producing 15 second video clips is 4x cheaper than previous options. The release expands the distribution of the Grok Imagine family beyond its primary interface.