xAI released Imagine Image 2.0 on August 7th, making it the new Quality Mode across grok.com/imagine and the company's mobile apps. The pitch is less about spectacle and more about control. According to xAI, the model follows a prompt down to the small details, renders legible typography, and handles editing tasks that older image models tended to fumble.
Those editing tools are the interesting part. Imagine 2.0 adds a magic wand for local changes, subject segmentation, background removal, and multi reference editing that accepts up to five input images at once. A smart resize feature reshapes a single image to any aspect ratio without an obvious crop. xAI also shipped a set of templates for common jobs like product shots, headshots, icons and game assets. On the Arena leaderboards, the company says the model ranks second in the world for both text to image generation and image editing. API access is listed as coming soon.
A busy day for xAI
The image launch arrived alongside Grok Build 1.0.0, the stable release of xAI's terminal based coding agent. That update is smaller in ambition, mostly cleaning up the command line experience: better session handling, clearer permission prompts that show the full script before it runs, tidier markdown tables, and steadier performance on long sessions.
Put together, the two releases show where xAI is pushing. Image generation and coding are the two areas where users notice quality gaps immediately, and where rivals have been moving fast. OpenAI reworked ChatGPT and opened its free tier this week, and Meta has been building out its own coding agent. Grok is now competing on the same two fronts at once.
Details on the release come from xAI's own changelog, collected by Releasebot. As with any first party benchmark, the leaderboard claims are worth checking against independent testing once the API opens up.
Commentarii · 0