Introduction
GPT Image 2 is an image generation model with exceptional text rendering, making it the strongest choice for any output where in-image text needs to be accurate and legible. Use it for mockups, posters, ad concepts, signage, social graphics, or any creative work where typography is part of the visual.
In addition to standard generation, GPT Image 2 supports image references and scene sketching to guide your output with visual inputs.
Spec information
| Unlimited generations in Explore Mode | Yes | ||
| Max reference images | 16 | ||
| Supported aspect ratios | 21:9 | ||
Credit cost by settings per image | |||
Resolution | Quality | ||
Low | Medium | High | |
1K & 2K | 1 | 5 | 20 |
4K | 2 | 11 | 41 |
Step 1 — Accessing GPT Image 2
GPT Image 2 is available in the Custom creation mode in Runway.
- Select Custom in the left sidebar
- At the top of the interface, confirm Image is selected — GPT Image 2 is an image model and will only appear when Image mode is active
- Open the model picker at the bottom of the interface and select GPT Image 2
Step 2 — Adding the inputs
GPT Image 2 accepts a text prompt and optional visual inputs.
Prompt
Describe the image you want to generate. Because GPT Image 2 has strong text rendering capabilities, you can include specific copy directly in your prompt — headlines, taglines, labels, or any text you want to appear in the output. Be direct about layout and content; the model interprets detailed prompts well.
Image references (optional)
Select add image references or drag and drop files to upload one or more images to guide the style, composition, or content of your generation.
Sketch (optional)
Select sketch a scene to draw a rough composition directly on the canvas, giving the model a visual layout to work from. If you've uploaded reference images, you can hover over the input to sketch directly on an image to guide placement, scale, and visually guide edits.
Step 3 — Generating with GPT Image 2
Before generating, review your settings at the bottom of the interface:
- Output count — set to 4 by default
- Aspect ratio — set to 16:9 by default
- Resolution — choose between 1K, 2K, or 4K
- Quality (within Settings
menu) — choose between Low, Medium, or High.
Credit costs are tied to both the selected Resolution and Quality settings.
Quality controls the level of detail rendered within your chosen resolution — higher quality produces more refined outputs at the same pixel count of the chosen Resolution. The default is High. Use Medium or Low to iterate more quickly or to conserve credits.
Once your prompt and settings are ready, select Generate.
Next steps
You've used GPT Image 2 to generate images with accurate in-image text. From here, you might explore apps to further work with your new images:
- Multi-Shot Video — Generate a multi-shot sequence to transform your image into a video
- Story Panels — Expand the world and story of your image with new perspectives