Grok Imagine Video 1.5 from xAI: A New Leap in AI-Powered Video Generation
xAI has officially released its new video generation model, Grok Imagine Video 1.5, from preview. This tool, which transforms images into dynamic video clips, is now available via API, as well as in the web service and mobile applications. The developers have focused on realism, correct physics of motion, and high rendering speed.
The model, known under the identifier grok-imagine-video-1.5, is available through the Imagine API. In parallel, xAI has launched an accelerated version — Video 1.5 Fast, which operates on the grok.com/imagine platform and in iOS and Android apps.
Key Improvements and Performance
The main speed increase came specifically in the Fast version. According to the developers, generation time has nearly halved — from over 40 seconds to approximately 25 seconds. Meanwhile, the clip length (6 seconds) and resolution (720p) remain unchanged.
Quality improvements have touched several important aspects. Sound accompaniment, background atmosphere, and character speech are now generated in a single pass, with all elements perfectly synchronized with the on-screen action. Character speech has become much clearer, and object movement is more stable throughout the clip. There are fewer deformations in frames, and weight and inertia look more realistic.
For developers, fine-tuning is available via the API: you can set a starting image, a text description of the motion, resolution, and duration.
Community Reaction and User Requests
Early users who tested the model for several weeks generally gave positive feedback, calling the team's work high-quality. However, there was a fly in the ointment. Attentive viewers noticed an unfortunate typo in the product name in the promotional material.
The main user requests concern the addition of voice references to create characters with consistent sound — a feature already implemented in competing services. Users also ask for the full versions of Video 1.5 and Imagine Pro to be opened not only via API but also 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.
My analysis: The release of Grok Imagine Video 1.5 from preview is a significant step for xAI in the generative video segment. The twofold speed increase and improved synchronization of sound and motion place the model on par with leading solutions. However, community requests for voice references and expanded access to full versions indicate that competition in this niche will only intensify. xAI needs to respond promptly to these wishes to avoid losing ground to more flexible platforms.