๐ UPDATE โ June 17, 2026
Elon Musk has just announced Grok Imagine 1.5, the latest major version of xAI's image generation model. The announcement dropped on June 17, 2026, accompanied by what appears to be a video showcase of the model's output quality โ suggesting a meaningful leap in visual fidelity or new capabilities beyond the Aurora-2 engine covered in our original article. With over 2 million views on the announcement post in under an hour, the update is generating significant buzz. Full feature details from xAI have not yet been officially documented. ๐
๐ UPDATE โ June 17, 2026
Grok has added a new integration with HeyGen's HyperFrames connector, letting users convert any Grok-generated answer into a polished video in minutes. This marks a significant expansion beyond Grok Imagine's native video generation โ now third-party tools can plug directly into Grok's output pipeline. To use it, simply add the HyperFrames connector via HeyGen and prompt Grok as usual; the connector handles the rest. This positions Grok as not just a generator, but a hub for multimedia workflows.
@grok ยท June 17, 2026Add the HyperFrames connector by @HeyGen to Grok and turn any answer into a video in minutes
![]()
๐ UPDATE โ June 17, 2026
xAI has launched Grok Imagine Video 1.5, a new image-to-video model now in wide release. It delivers sharper realism and improved physics, with 720p videos rendering in approximately 25 seconds โ down from 40+ seconds in the previous model. The model is generally available via API, and a Video 1.5 Fast variant has also rolled out to consumers with better quality and dramatically reduced wait times. Elon Musk confirmed the wide release on X shortly after xAI's official announcement.
![]()
via @xai on X
Elon Musk just told his 200M+ followers to try the latest Grok Imagine โ and if you haven't checked in recently, you've missed a lot. xAI's image and video generation tool has undergone a rapid series of upgrades since early 2026, quietly becoming one of the most capable AI creative platforms available. Here's a rundown of the most significant changes.

What's New in Grok Imagine
1. Grok Imagine 1.5 Preview โ Now Live
xAI officially announced the Grok Imagine 1.5 Preview on June 3, 2026, with the release formally noted in xAI's own release notes on June 11. This is the version Musk is pointing users toward now. It builds on the Aurora-2 engine foundation introduced earlier this year and represents the current state-of-the-art for the platform's image generation capabilities.
2. Full Video Generation Has Arrived
This is the headline feature. On June 5, Musk shared a demonstration confirming that Grok Imagine has expanded well beyond still images into full video generation โ a major leap from its image-only origins. The underlying Grok Imagine 1.0 model, powered by the Aurora-2 engine, generates clips up to 10โ30 seconds at 4-megapixel native resolution. At 720p, the API prices video generation at $0.07 per second (a 10-second clip runs approximately $0.70). A formal announcement through official xAI release notes is still pending, but the capability is live.
3. Quality Mode vs. Speed Mode
Introduced around May 22, Quality Mode is one of the most practically useful additions for anyone using Grok Imagine for real creative work. Compared to Speed Mode, it delivers sharper textures, more accurate lighting, and โ critically โ dramatically improved text rendering in generated images. If you've been frustrated by garbled text in AI-generated visuals, Quality Mode is worth switching to immediately.
4. Aurora-2 Engine and 4-Megapixel Output
As of April 22, Grok Imagine 1.0 runs on xAI's proprietary Aurora-2 engine, which pushes native 4-megapixel resolution for images and includes integrated "Voice Doctor" audio synthesis for audio-visual synchronization in video outputs. The audio is described as context-aware โ meaning the system generates sound that matches the visual content rather than applying a generic soundtrack.
5. "Extend from Frame" for Continuous Sequences
Introduced on March 2, this feature lets users chain video clips together, with each new generation picking up precisely where the previous one ended. Each clip can run up to 15 seconds, making it possible to build longer continuous visual sequences without jarring cuts. It's particularly useful for storytelling or product demonstration workflows where continuity matters.
6. Game Asset Prototyping Workflow
In May, xAI officially demonstrated a game asset prototyping pipeline that chains Grok Imagine's image generation and image-to-video models together to produce animated game characters. This signals that xAI is positioning Grok Imagine not just as a consumer novelty but as a legitimate tool for professional creative and development workflows.
7. API Now Open โ With Transparent Pricing
The Grok Imagine API launched on January 28, 2026, giving developers access to the full creative pipeline. Current pricing: text-to-image at $0.02 per image, image editing at $0.022 per image, and video generation at $0.05/second (480p) or $0.07/second (720p). For teams building AI-assisted creative tools, this is a straightforward entry point into xAI's generation stack.
The Bottom Line
Grok Imagine has moved fast โ from a basic image generator to a full video production platform with professional-grade features in under six months. Musk's nudge to "try the latest" isn't just a casual post; the 1.5 Preview represents a genuinely different product from what most users last tested. If you haven't opened Grok Imagine since early this year, the gap between then and now is significant enough to warrant a fresh look.

Sarah focuses on Tesla Energy, SpaceX missions, and the broader Musk AI portfolio. Former data analyst in clean energy. Based in San Francisco.
Sources verified at publish time. Spotted an inaccuracy? Email editorial@basenor.com.







