AI competition is becoming even more intense. OpenAI and Google are already rolling out increasingly sophisticated text-to-video generation models, but Elon Musk-owned xAI is looking to stay in the competition. The company just announced its new Grok Imagine Video 1.5 model and believes creators not only want great-looking videos, but faster results and more creative control. According to xAI, Grok Imagine Video 1.5 is the “best image-to-video model ever,” offering better motion, better physics, better audio, and faster speeds.
The announcement comes as AI companies are actively racing to turn simple prompts and images into lifelike videos. However, generating compelling videos remains one of the most difficult challenges in AI.
Better video with built-in sound
According to xAI, the new model is improved in all aspects important for real-world creative work. Compared to previous versions, users get better sound effects, atmosphere, and dialogue, all generated in the same pass and synchronized with the action. The company also says that audio is now clearer and syncs better with video.
This is important because audio is often treated as a separate layer in AI-generated videos. xAI aims to create visuals and sounds together so that the final output feels more natural.
Solving one of the biggest problems with AI video
Physics is another area where AI models have long struggled. Users often complain about videos where objects bend unnaturally, characters float awkwardly, and movements look unrealistic. According to xAI, Grok Imagine Video 1.5 addresses some of these issues by reducing distortion and delivering more reliable weight and momentum.
Whether that translates into significantly better video ultimately depends on user response, but the company has made physics one of the model’s biggest selling points.
Accelerating video generation
xAI is also focused on speed. The company says Grok Imagine Video 1.5 Fast nearly doubles the production speed. You can now create a 6-second video at 720p resolution in about 25 seconds, compared to over 40 seconds with the previous model.
In the rapidly changing AI market, shaving a few seconds off generation time can make a big difference, especially for creators who generate multiple versions before deciding on the final output.
New tools for creators
Alongside Imagine Video 1.5, xAI is rolling out several new features in the coming days to improve users’ creative workflows. Users can organize their work into projects that appear in the left sidebar. The company is also introducing parallel agents, which will allow users to run multiple prompts at once instead of waiting for one generation to finish and start another.
Another added feature is search. Users can find images and videos they’ve created by searching their library instead of scrolling through older works.
Imagine Video 1.5 is now in preview and generally available in the xAI API as “grok-imagine-video-1.5”. Users can provide a starting image, describe the movement, and choose the video resolution and length.
xAI has also rolled out Video 1.5 Fast at grok.com/imagine and its iOS and Android apps.
– end
