xai / grok-imagine-image

## Overview Grok Imagine Image is an image generation and image editing model from xAI. It’s part of the Grok Imagine stack, built for prompt-following and fast visual iteration. Use it to generate new images or transform an existing photo with text instructions. ## What you can build - Text-to-image scenes for storyboards, pitch decks, and concept art - Poster and social graphics that need readable on-image text - Product mockups like packaging, labels, and simple ad creatives - Photo edits like wardrobe color changes, background swaps, and weather effects - Style changes such as watercolor, anime, pencil sketch, or pop-art looks - Batch variations for A/B testing compositions and lighting ## Inputs - A required written instruction that describes the image you want, or the edit you want - An optional single source image for editing - Max file size: 20 MiB - Supported formats: JPG, JPEG, PNG - Only 1 image is allowed - A requested number of output images to generate (1 to 10) - A target aspect ratio for text-to-image generations - If you provide a source image, the output keeps the same aspect ratio as the input image - An output resolution setting - 1K for standard output - 2K for higher-resolution output ## Outputs The model returns 1 to 10 generated or edited images. Outputs are delivered as image files, typically as downloadable assets or links provided by the platform. ## Recommended settings - Quick exploration (fast prompt iteration) - Samples: 1 to 3 - Resolution: 1K - Aspect ratio: 1:1 for square feeds, or 16:9 for wide banners - Final assets for detailed images - Samples: 1 to 2 - Resolution: 2K - Use a focused prompt with specific details and fewer competing style cues - Clean image edits (keeping the original structure) - Upload a source image - Describe only the changes you want - Keep samples low if you need consistent results across revisions ## Limitations - You can request up to 10 images per run. - Output resolution is limited to 1K or 2K. - Image editing accepts only 1 source image in this Wiro setup. - Generated image links may be temporary, depending on how results are delivered. - Outputs are subject to content moderation. ## Safety & compliance This model applies content moderation to prompts and outputs. Don’t use it to create or edit content that violates laws or platform rules, including non-consensual nudity, sexual content involving minors, or harassment. Only upload images you have rights to use, and get consent when editing real people’s photos.