Canonical page: https://mixerai.org/en/knowledge/how-to/text-to-image/

[Mixer AI](https://mixerai.org/) › [Knowledge Base](https://mixerai.org/en/knowledge/) › [How to](https://mixerai.org/en/knowledge/how-to/) › How to create an image from a description

# How to create an image from a description

A useful prompt explains what belongs in the frame and how to show it. Start with the main subject, then composition, light and medium. A pile of impressive adjectives does not replace a clear scene.

[Try it on Mixer AI](https://mixerai.org/?dl=kb_text-to-image#/canvas)

## Step by step

Name the subject and setting. Replace 'stylish ad' with a glass bottle on pale stone, soft side lighting and a quiet background.

Choose an aspect ratio for the placement. Leave room for banner text and avoid pushing a portrait's head against the edge.

Generate a test and identify its main issue. Adjust composition, light or materials one at a time while preserving what works.

## Which model to use

[

GPT Image 2.5

Flare for exploring ideas, Sunburst for demanding edits, with a shared image creation workflow.

](https://mixerai.org/en/knowledge/image/gpt-image-2-5/)[

Midjourney

Explore a visual direction through lighting, material, composition and mood.

](https://mixerai.org/en/knowledge/image/midjourney/)[

Z-Image

A new image from a short description, for concepts, illustrations and first visual experiments.

](https://mixerai.org/en/knowledge/image/z-image/)

## FAQ

Must I specify a camera brand?

No. Framing, light direction and depth of field often explain the desired look more clearly.

How do I add exact words?

Put them in quotes and specify placement. Verify or add typography in an editor when precision matters.

How do I preserve a particular product?

Text may invent a similar object. Use a photo-reference model for an accurate source product.

[Try it on Mixer AI](https://mixerai.org/?dl=kb_text-to-image#/canvas)