GPT Image 2 – High-Fidelity AI Image Generator

GPT Image 2 is OpenAI's image generation and editing model. It produces high-fidelity images from text prompts and can edit or compose scenes using up to 16 reference photos.

GPT Image 2 favors accurate typography, clean vector-like illustrations, and strong subject consistency across edits. Nano Banana 2 emphasizes photorealism and Google Search-grounded generation. Pick the one that matches your style.

Yes. It is one of the strongest models for rendering legible, correctly spelled text — useful for posters, infographics, storefronts, and UI mockups.

Up to 16 reference images. When at least one image is provided, the request runs in image-to-image mode and the model preserves subject identity while applying your prompt.

At 1K, Auto and these aspect ratios are supported: 1:1, 3:2, 2:3, 4:3, 3:4, 5:4, 4:5, 16:9, 9:16, 2:1, 1:2, 3:1, 1:3, 21:9, 9:21. Auto lets the model choose the ratio. Adding references does not change your selection.

Each image costs 2 credits at 1K, 8 at 2K, or 13 at 4K. The price is the same for generation and editing, regardless of reference count. Create up to five images at once; the total updates before you submit. Auto ratio and background selection are available only at 1K. Unsupported ratios are disabled for each resolution.

Be specific about subject, setting, style, and composition. For text-in-image, wrap the exact text in quotes. If you upload a reference, describe what to keep and what to change.