Back to models
OpenAI
Image modelHotGPT Image 2
GPT Image 2 is OpenAI's image generation and editing model. It creates images from text and accepts existing images as input for instruction-based changes. It suits work that must balance subject details, text within the image, and composition, such as product scenes, brand proposals, advertising drafts, and edits that preserve a person's features. Describe the desired final image, then specify style, viewpoint, aspect ratio, and any required text. For editing, identify the changes and which backgrounds, objects, or details must stay. It offers low, medium, and high quality settings and sizes from 1K to 4K. Editing can use multiple reference images, allowing the same workflow to cover different layouts and assets. The model helps produce usable drafts, but commercial delivery still requires checks of text, logos, people's appearance, and product details. For testing multiple versions, keep the prompt and image dimensions fixed when comparing results, so differences in settings are not mistaken for differences in model capability. Content teams can use it for a continuous workflow from ideas and generation to targeted revisions.
- High-Quality Generation
- Detailed Editing
Key features
Integrated generation and editing
Create images from text or edit existing content with multiple input images and instructions.
Image text and composition
Suits assets with headings, product information, or specified layouts; proofreading is still required before final use.
Quality and size options
Offers three quality levels and 1K, 2K, and 4K output to match the intended use.
