Elon Musk’s xAI has launched Imagine Image 2.0, a new generation of its image creation and editing technology, as competition among AI companies shifts increasingly towards controllable, production-ready visual generation.
The model is now generally available as the new Quality Mode on Grok’s Imagine platform and through its iOS and Android applications. API access is expected to follow.
Unlike AI image generators primarily designed around creating visuals from prompts, Imagine Image 2.0 places greater emphasis on editing and creative control. xAI says the model has improved instruction following and can better maintain typography, layouts and visual consistency across generations.
One of the biggest additions is region-level editing, allowing users to select specific portions of an image and describe the changes they want without regenerating the entire visual. Users can also provide multiple reference images and combine elements such as people, products, objects and environments into a new composition.
The release also introduces templates for common creative workflows, including removing objects, changing backgrounds, modifying outfits and adjusting elements within existing images.
Early benchmark results suggest xAI has become a serious challenger in the increasingly competitive AI image market. Imagine Image 2.0’s faster variant ranked second on both the Image Edit Arena and Text-to-Image Arena around its launch, behind OpenAI’s GPT-Image-2.
The launch strengthens Grok’s transformation from an AI chatbot into a broader multimodal creation platform. With OpenAI, Google, Adobe and other AI companies investing heavily in image generation and editing, differentiation is increasingly moving beyond image quality towards control, consistency and usability in real creative workflows.
For marketers and creative teams, that shift could ultimately prove more significant than simply generating better-looking AI images.