An image from a description
Describe the subject, setting, light and style, pick one of 5 aspect ratios, and Grok Image creates the picture. Handy for turning an idea into an image: an illustrated cover, a social media visual, a mood photo.
Image model · xAI
Grok Image is Deepnia’s name for Grok Imagine, the image model from xAI, the company behind the Grok assistant. It creates an image from text in one of 5 aspect ratios (1:1, 2:3, 3:2, 16:9, 9:16), or edits a photo you add by following a plain-language instruction; the edited image keeps your photo’s aspect ratio.
The image tool opens with Grok Image already selected. The exact cost shows before every creation; you sign up when you generate.
Describe the subject, setting, light and style, pick one of 5 aspect ratios, and Grok Image creates the picture. Handy for turning an idea into an image: an illustrated cover, a social media visual, a mood photo.
Add a photo and describe the change in plain language: a new backdrop, another season, different light, an object added or removed. xAI documents this instruction-based editing for Grok Imagine.
When you edit, the result automatically keeps the original photo’s aspect ratio, whether it is vertical, square or horizontal: nothing to set.
xAI describes step-by-step editing: you start again from the image you got to ask for the next change. On Deepnia, add the generated image as a reference from your library, then write the new instruction.
Up to 5,000 characters to spell out the subject, composition, light and style, whether you create an image or edit a photo.
The “Try Grok Image” button opens Deepnia’s image generator with the model already selected.
Without an image, you create from text. Use “Add an image” under “References” (1 photo at most) and the tool switches to “Edit mode”.
Describe the image to create, or the change to make, in up to 5,000 characters.
Without a photo, pick the aspect ratio (1:1, 2:3, 3:2, 16:9, 9:16). Check the cost shown, then click “Generate”, or “Edit” when you are editing a photo.
Copy a prompt, open the tool and adapt it to your project.
A superhero comic book cover: a young heroine in an orange and midnight-blue suit glides above a big city at dusk, her cape snapping in the wind, neon signs lighting up below. Bold inked lines, saturated colors, 1980s comic halftone texture. No text or title.
From your photo, replace the background with a white marble counter in a bright kitchen, with a blurred green plant behind. Keep the product exactly the same: shape, colors, label and reflections. Add a soft, natural shadow under the product.
Travel photo of a small fishing harbor at sunrise: colorful boats on calm water, white houses with blue shutters on the hillside, light mist, golden reflections. Travel photography style, wide angle, natural colors.
Square visual announcing a rooftop birthday party: string lights, gold and navy balloons, a table of colorful cocktails, the city lit up behind. Festive, elegant mood, with empty space at the top to add text later; no text in the image.
From your photo, turn the scene into a winter day: fresh snow on the roofs and sidewalks, falling snowflakes, soft cold light. Keep the composition, the buildings and the people exactly where they are.
Realistic photo of a fitness coach cheering on a group during an outdoor session at sunrise in a city park: golden light, smiles, upbeat energy. Advertising photo style, shallow depth of field, vertical framing for a story.
Deepnia’s other image models, to pick the right one for your project.
| Model | Resolution | Aspect ratios | References | Best for |
|---|---|---|---|---|
| Grok ImageThis model | Automatic | 5 ratios | Up to 1 | An image from text, or your photo edited with one sentence. |
| Nano Banana 2 Lite | 1K | 14 ratios | Up to 10 | Deepnia’s default image model, to create or edit with several references. |
| Nano Banana 2 | 1K–4K | 14 ratios | Up to 14 | Images in 1K, 2K or 4K, delivered as PNG or JPG. |
| GPT Image 2 | 1K–4K | 8 ratios | Up to 16 | Text-rich images: infographics, posters and graphic design. |
| Seedream v5.0 Lite | Automatic | 7 ratios | Up to 10 | 2K images, created or edited from several references. |
| Z-Image | Automatic | 5 ratios | None | Photorealistic images from a simple description. |
It is Grok Imagine, xAI’s image model, on Deepnia. It creates an image from text or edits a photo you add, following a plain-language instruction.
Yes: add a photo under “References” and describe the change. The tool switches to “Edit mode” and the result keeps your photo’s aspect ratio.
1 photo, the one you are editing: add it under “References” (JPG, PNG or WebP, up to 10 MB) and the tool switches to “Edit mode”. To combine several images in one creation, choose Nano Banana 2 Lite.
1:1, 2:3, 3:2, 16:9, 9:16, with 9:16 by default, for an image created from text. With a photo, the result automatically keeps the photo’s aspect ratio.
Make one change at a time: add the image you got as a reference from your library, then write the next instruction. xAI describes this step-by-step editing for Grok Imagine.
Grok Image to turn an idea into a picture or to edit a photo with one sentence. GPT Image 2 for text-rich visuals such as infographics, posters and mockups.
1 image per generation on Deepnia. Click “Regenerate” to get another version from the same request.
Open Deepnia’s video tool and choose Grok Video: it creates a video from text or from an image.
The image tool opens with Grok Image already selected. The exact cost shows before every creation; you sign up when you generate.