Image generation creates images from text prompts or edits existing images. Moxus AI routes supported image models through the same account, API key, quota, and usage log system.
You can first choose an image generation model in the web Conversation page and test your prompt there. When you are ready to integrate it into a product, use the API or SDK from your server-side code.
Capabilities
- Text to image: generate an image from a written prompt.
- Image to image: adjust style, content, or composition based on a reference image.
- Local edits: modify part of an image when the selected model supports it.
- Creative batches: generate multiple options for the same concept.
Image models differ in size, quality, number of outputs, reference image support, async behavior, and pricing. Check model details before submitting production requests.
Prompt structure
- Subject: what to generate.
- Style: photo, illustration, product render, poster, and so on.
- Composition: close-up, top-down, centered, or with negative space.
- Constraints: colors, aspect ratio, and elements to avoid.
- Use: avatar, cover, app asset, or marketing visual.
API example
The cURL, Python, and Node.js examples for image generation are maintained in Vision and image generation. They use grok-imagine-image.
Next steps
