DALL·E 3: Describe your image and create

Describe your idea, and DALL·E 3 creates illustrations, character images, structured scenes, story visuals, and educational visuals with clear composition.

ToolAI Text to Image

ModelDALL·E 3

StyleAuto

Aspect Ratio1:1

Use casesEducational illustrations

What can you do with GPT Image?

Six core capabilities for all your image creation needs

Text-to-image

Describe the scene in words and AI paints it for you. No drawing skills needed to create professional images. Example: "A dog in sunglasses on a beach vacation"

Image-to-image

Upload a reference image and tell AI the style you want. It will redesign it for you. Example: Upload a photo + "Change to oil painting style"

Local edits

Edit only the part you want to change. Leave everything else as is. Example: Change the outfit in a photo

Multi-version generation

Generate multiple images at once and pick the best one. Example: Same character in 4 different poses

Business assets

Quickly create posters, covers, illustrations, product shots, and more. Save time and effort. Example: Generate e-commerce store hero images

Style consistency

Keep multiple images in a consistent style, like having a professional design team for your brand visuals. Example: 10 unified social media posts

Three steps to get started

Follow along—no prior experience needed

Choose mode

Decide what you need: generate from text or edit an existing image? Three modes Text-to-image Upload image to modify Local edit

Write your prompt

Describe what you want in plain language, like chatting with a friend. Universal formula Subject + scene/style + Key details + Elements to exclude (optional)

Set parameters

Pick image size, quality, then hit generate. Common settings Size: 1:1 / 16:9 / 9:16 Quality: Low / Medium / High Count: 1 / 2 / 4 images

What does it look like?

See what others have created with GPT Image

Text-to-image: From zero to one
No source material needed—describe the scene and AI draws it. Great for quick images when you have no assets.

Prompt example

A cat in a suit working in an office, city view through the window, sunlight streaming in

Case description

Text-to-image: From zero to one No source material needed—describe the scene and AI draws it. Great for quick images when you have no assets.

Can't write prompts? Just copy

These templates are proven to work—tweak and use

Product shot

[Product name], pure white background, [material], professional product photography, [angle], 8K quality, clean background, e-commerce hero style 💡 Tips More realistic→add "natural light", "shadow" | More premium→add "minimalist", "luxury"

Use template

Poster

[Theme], [style: modern/vintage/tech], [color palette], commercial poster, clear hierarchy, strong visual impact, for [use case] 💡 Tips More eye-catching→add "high contrast" | More professional→add "whitespace", "minimalist"

Use template

Social cover

[Content topic], [style: fresh/bold/cute], social media cover, unified colors, 16:9 ratio, clear focal point 💡 Tips Xiaohongshu→add "bright" | WeChat→add "refined" | Douyin→add "dynamic"

Use template

Illustration

[Scene], [style: flat/gradient/hand-drawn], [colors], vector style, clean and simple, modern, for [use case] 💡 Tips Children's book→add "cute" | Commercial→add "professional" | Article→add "minimal"

Use template

GPT Image works better here

It's not just the model—the platform makes it smoother

🌐 Multi-model hub

Switch between models for the same task and compare results to find the best fit

💾 Workflow management

Save prompts, reuse templates, batch generate, version control—create more efficiently

🔄 Smooth iteration

Full loop from idea→draft→final. Not satisfied? Edit until you are.

FAQ

GPT Image mainly does three things: 1) Generate images from text; 2) Change style with a reference image; 3) Edit specific parts. In short: generate, edit, retouch.