Second in the world: Musk SpaceXAI launches Imagine Image 2.0 to strengthen AI image generation/editing capabilities

IT House reported on August 8 that SpaceXAI, a company owned by Musk, released a blog post today (August 8), announcing the launch of the Imagine Image 2.0 model, which is now available as a new quality mode (Quality Mode) in the Grok.com web version, iOS and Android applications. In terms of generation effects, the official website states that the Imagine Image 2.0 model follows user instructions and can handle detailed requirements; in complex visual content, it will plan text layout and layout structure to keep multiple parts of the picture consistent and improve the clarity of small font text. At the same time, the model can retain user-input entities and settings across multiple rounds of generation and editing. IT Home attaches the relevant schematic diagram as follows: Imagine Image 2.0 focuses on enhancing the image editing capabilities. SpaceXAI said that in real work scenarios, multiple rounds of modifications are usually required, and the first generated picture is rarely the final product. Therefore, a number of tools for local modification and material reuse are added to the model: The magic wand tool only edits the area pointed by the user while keeping other areas unchanged; the segmentation function can select the precise area in the image and then modify it for that area; The background removal function can export the subject with a transparent background for easy inclusion in other works; multi-reference image editing supports the input of up to 5 images in a single generation, reducing the need for manual synthesis. Intelligent size adjustment, after the user selects the target ratio, the model will complete the frame of the picture, allowing the same picture to adapt to different size requirements. In terms of performance, Image 2.0 is geared towards photography, design and illustration scene training. According to the Arena Text-to-Image (Arena Text-to-Image) and Image Edit (image editing) ranking data cited by xAI, as of August 7, 2026, Image 2.0 ranks second in the world in both text-to-image generation and image editing capabilities, second only to OpenAI’s GPT Ima