xAI launches Grok Imagine Video 1.5: a new level of AI video generation — detailed breakdown
xAI has officially taken its new model, Grok Imagine Video 1.5, out of preview mode. This tool is an image-based video generator, with developers focusing on realism, correct motion physics, and high rendering speed. The model is already available via API, and users have gained access to an accelerated version of the technology.
What the new model can do
Grok Imagine Video 1.5 is now available in the Imagine API under the identifier grok-imagine-video-1.5. Simultaneously, the team has deployed a Video 1.5 Fast version, which operates on the web service grok.com/imagine, as well as in iOS and Android apps.
| Parameter | Previous model | New Fast version |
| Generation time | Over 40 seconds | About 25 seconds |
| Clip length | 6 seconds | 6 seconds |
| Resolution | 720p | 720p |
The main speed increase applies specifically to the Fast version: xAI claims nearly double the performance.
Quality improvements have touched several important areas. Sound accompaniment, background atmosphere, and character dialogue are now created in a single pass. All elements are perfectly synchronized with the on-screen action, and character speech has become much clearer.
Object movement remains more stable throughout the entire clip. There is less deformation in frames, and weight and inertia look more realistic. Through the API, developers can set a starting image, a text description of movement, resolution, and duration.
Alongside the model, xAI is gradually introducing useful tools for the workflow. The interface now includes projects in the sidebar, parallel execution of multiple generations, and convenient search through a personal material library.
How users reacted
Part of the audience has been testing the model for several weeks. Early users evaluated the result positively, calling the team's work high-quality. But there is a fly in the ointment. For example, in promotional material, attentive viewers noticed an unfortunate typo in the product name.
Several commenters formulated their wishes. Among the main requests is the addition of voice references. The feature is necessary for creating characters with consistent sound, as in competing services.
Additionally, authors are asking for the full versions of Video 1.5 and Imagine Pro to be opened not only via API. Users want to see them in the app and on the website, even with limitations. Separately, there were complaints about abrupt changes in the interface and grievances about increased content moderation for adult users.
Analytical commentary from Cryptalist: Taking Grok Imagine Video 1.5 out of preview mode is a significant step for xAI, especially against the backdrop of fierce competition in the generative video segment. The twofold acceleration in generation and improved motion physics are exactly what content creators need. However, as feedback shows, users are already expecting not just speed, but more subtle settings, like voice references. If xAI can quickly implement these features, the project has every chance to carve out its niche. For now, the market remains dominated by giants like OpenAI and Runway, and falling behind in the race for quality could be costly.