Create AI images from text

Describe your subject, scene, and style, choose an available model, and generate an image from your prompt. No source image is required.

Start with a written idea

Describe the subject, composition, lighting, and details in your prompt.

Choose a model and output settings

Select an available model and use the aspect ratio and output options supported by that model.

Generate, review, and refine

Review the output and refine your prompt to explore alternatives. Beta access depends on available allowances and capacity.

How to use Text to Image

Prompt

Editorial studio portrait of a stylish young woman posed against a vivid red set with a keyhole-shaped opening revealing a blush pink backdrop. She has sleek lavender hair, a sculpted satin mini dress with floral accents, lace tights, and a white bow with a heart-tipped arrow, lit with crisp high-contrast studio light for a polished fashion-magazine look.

1. Write your prompt

Write a description of the image you want to create. You do not need to upload an image.

Settings

Choose an available model and its supported output options.

2. Choose generation settings

Choose an available model and its supported output options.

Result

3. Generate and review

Submit the request, review available results, and revise your prompt if you want another direction.

Text to Image FAQs

Do I need to upload a source image?

No. Text to Image starts from a written prompt. To edit an existing image, choose the Image to Image workflow instead.

How can I write a clearer prompt?

Specify the subject, scene, composition, and style, then make focused changes when refining your prompt.

Can I try different styles?

Yes. Revise your prompt or select another available model. Features and access depend on the model and beta allowances.

Start creating images from text

Enter a prompt and choose an available model to explore your next image.

Text to Image

Examples