SAN FRANCISCO — Elon Musk marked the U.S. Independence Day weekend with a two-word progress note that lands squarely in xAI's product roadmap. On July 5, the founder posted 'Done with Grok Imagine,' signaling that the core development cycle for Grok's built-in image and video generator has wrapped up. It is a milestone that moves xAI's flagship assistant well beyond text and into fully multimodal territory.
From Text Bot to Visual Engine
Grok Imagine has been in the works since xAI first teased the feature in August 2025, followed by an initial beta in October of that year. Under the hood, it leans on xAI's proprietary Aurora model to turn written prompts into images, with support for seven aspect ratios and a broad range of visual styles. xAI paid particular attention to accurate text rendering inside generated images, historically a weak spot for rival generators, and the tool now handles short-form video in addition to stills.
The feature has been rolling out through xAI's premium tiers, reaching SuperGrok and Premium+ subscribers on the Grok iOS app first, with Android access extended to heavier users. Crucially, it is also wired directly into the X app, giving Grok Imagine a built-in distribution channel measured in hundreds of millions of potential users. That pairing of a finished creative engine with an enormous audience is exactly the kind of flywheel xAI has been building toward, much as it did when it rolled out its no-code voice agent builder for Grok.





