About this model
Grok Imagine High Quality is the image-editing (inpainting) member of xAI's Grok Imagine creative suite, released in May 2026. It applies prompt-driven edits and inpainting to existing images, complementing the quality text-to-image path xAI documents under its image generation API. The model sits inside the broader Grok Imagine bundle, which xAI describes as a unified set of APIs for end-to-end creative workflows spanning image, video, and audio generation.
Within its own family, it succeeds the original Grok Imagine inpainting model from January 2026, moving editing onto xAI's newer "quality" image stack. It is the editing counterpart to the Grok Imagine High Quality (SOTA) text-to-image model, and to the earlier Grok Imagine image generator. xAI's docs note that the older grok-imagine-image-pro slug is being deprecated in favor of the quality image model, signaling the family's consolidation around this generation.
xAI documents image editing support through its Imagine image models, and states that Grok Imagine focuses on instruction following, which matters for precise, region-targeted edits.
Public, primary benchmark figures specific to this editing model were not available at the time of writing, so this description stays to documented capabilities. The model is reachable through xAI's unified Grok Imagine APIs. It belongs to the same wave as siblings like Grok 4.3 and Grok Imagine 1.5 Private.
This About section is AI-generated from public sources (Claude Opus 4.8), with no human editing. It may contain inaccuracies — verify critical details against the sources listed above.
Data sources: Venice API · HuggingFace · Wikipedia — enrichment updated 4d ago