Grok Imagine Video 1.5: xAI promises more realism and speed

xAI launches Grok Imagine Video 1.5, an image-to-video model on its API and grok.com featuring synchronized audio and a Fast version that doubles render speed.

xAI is advancing its video generation with Grok Imagine Video 1.5, its image-to-video model. From a starting image and a motion description, it produces more realistic shots, according to the company, with more consistent physics and movements throughout a clip's duration and fewer distortions. Audio is generated in the same pass as the image: sound effects, ambient sounds, and dialogues aligned with the action, with improved speech synchronization.

The full model is exiting the preview phase and becomes accessible in the xAI API under the identifier grok-imagine-video-1.5, with selectable resolution and duration. For the general public, a Video 1.5 Fast variant is coming to grok.com and the iOS and Android applications: according to xAI, it almost doubles the generation speed, with a six-second 720p clip being rendered in about twenty seconds compared to over forty previously. Progressively, work organization tools are being added: Projects organized in a sidebar, the ability to start multiple agents in parallel on the same project, and a search function within one's creation library.