OpenAI
GPT Image 2
GPT Image 2 in DesignCat: text to image, image to image. Explore supported settings, input requirements and model costs.
Choose the operation, then the settings.
GPT Image 2 supports 2 operations in this catalog. Settings belong to an operation: a reference limit or duration from one route does not automatically apply to another.
Use it in DesignCat
Mobile availability is described separately on the mobile page.
Text to image
Output category: image. Cross-field and source-media requirements are validated when you submit.
| Input or setting | Support and limits |
|---|---|
| Prompt · required | Required input |
| Num Images | minimum 1 · maximum 4 · default: 1 |
| Aspect Ratio | 16:9, 9:16, 4:3, 3:4, 1:1 · default: 16:9 |
| Resolution | 1K, 2K, 4K · default: 1K |
| Custom Resolution Width | maximum 3840 |
| Custom Resolution Height | maximum 3840 |
Image to image
Output category: image. Cross-field and source-media requirements are validated when you submit.
| Input or setting | Support and limits |
|---|---|
| Prompt · required | Required input |
| Num Images | minimum 1 · maximum 4 · default: 1 |
| Aspect Ratio | 16:9, 9:16, 4:3, 3:4, 1:1 · default: 16:9 |
| Resolution | 1K, 2K, 4K · default: 1K |
| Custom Resolution Width | maximum 3840 |
| Custom Resolution Height | maximum 3840 |
| reference · required | Required input |
| reference.image ids · required | up to 16 items |
| Mask Image Id | Optional input |
Quality and cost
Generation costs depend on the operation and settings. The application displays the applicable estimate before submission. Published comparison costs include their own settings and assumptions.
Continue with a workflow
Choose a workflow whose inputs and model support match your task.
Catalog source updated 2026-10-17. Model key: gpt-image-2. Public data and settings.