xAI released Grok Imagine Image 2.0 on August 7 as the new Quality Mode on grok.com/imagine and in the Grok iOS and Android apps. The model now sits second in the world on both major image leaderboards, behind only OpenAI's GPT-Image-2.
On the Arena leaderboards as of August 7, the faster “low” variant of the model scores an Elo of 1,320 in the Text-to-Image Arena, trailing GPT-Image-2 at 1,380. In the Image Edit Arena it hits 1,439, again behind GPT-Image-2 at 1,463. The model is listed under the SpaceXAI name on Arena.
The core pitch is editing. Imagine Image 2.0 ships a tool called Magic Wand that modifies only the selected region of an image, plus a segmentation feature for precise areas and a background removal tool that exports subjects on a transparent background. Multi-Ref Editing combines up to five input images into a single generation, and Smart Resize converts an existing image to any aspect ratio while the model fills in the extra space.
xAI trained the model for fidelity across photography, design, and illustration, with editing treated as a first-class capability. The company says it is designed to follow instructions with fine-grained accuracy, keep typography and layout clean in complex visuals, and stay consistent across multiple generations.
The rankings are the company's own figures drawn from the public Arena leaderboards, not an independent evaluation. The announcement also notes that API access for Imagine Image 2.0 is coming soon.
