Grok Imagine Image
xAI's image model: bold, photoreal images that follow your prompt.
What it's best at.
- Among the best-ranked text-to-image models on public leaderboards.
- Photorealism and bold styles that stick closely to your prompt.
- Edits guided by up to 5 reference images.
Specs.
- Resolutions
- 1K
- Formats
- 1:1 · 16:9 · 9:16 · 3:2 · 2:3
- Reference images
- Up to 5
- Plans
- Creator, Plus, Ultra and Teams
What it costs.
Credits per generation. You see the price before you start, and failed generations are refunded.
| Resolutions | Standard |
|---|---|
| 1K | 10 credits |
Plus gives you 6,000 credits a month.
Prompts to try.
Copy one as it is, or make it yours.
A night market in the rain, steam rising from food stalls, neon reflections on wet stone, cinematic photo.
A bold screen-print poster of a tiger leaping through jungle leaves, orange and teal, grainy texture.
Put the dog from my photo in tiny knight's armour on a castle drawbridge, same dog, golden hour.
Questions.
From 10 credits per image. You see the exact price before you generate, and if a generation fails, your credits come back.
Creator, Plus, Ultra and Teams. Every plan includes commercial use, with no watermark.
No. Grok Imagine Image runs inside Aventada with your credits. No extra account, no API key.
Yes. Connect Aventada to Claude or any MCP app once, then ask for a generation in the chat. It uses your Aventada credits.
Other models.
All models- VideoGrok ImaginexAI's quick, playful video model, with sound, included in Starter.
- VideoGrok Imagine 1.5xAI's newest video model: your picture, brought to life with sound.
- ImageNano Banana ProCreate and edit images with the same face, every time.
- ImageNano Banana 2Fast, consistent images and edits, included in Starter.
- ImageNano Banana 2.1The newest Nano Banana: better images for fewer credits, included in Starter.
- ImageGPT Image 2Sharp text, posters and clean design.
Use cases