GPT Image 2

Generate high-quality images from text or references with OpenAI's GPT Image 2 — 1K, 2K, and 4K output with 20,000-character prompts.

Resolution: 1K, 2K, 4K
Aspect Ratios: 1:1, 16:9, 9:16, 4:3, 3:4, auto

Create with GPT Image 2 View Pricing

GPT Image 2 sample — Y2K bubblegum magazine cover with bold pink typography

What makes GPT Image 2 unique

Supports both text-to-image generation and image-to-image workflows for creating new visuals or transforming existing ones.

Generates images at 1K, 2K, and 4K resolutions to support everything from fast concepts to production-ready assets.

Offers flexible aspect ratios including auto, 1:1, 9:16, 16:9, 4:3, and 3:4 for social, web, and marketing formats.

Accepts up to 16 reference images for image-to-image generation, enabling more controlled visual direction.

Supports prompts up to 20,000 characters, giving teams room for detailed creative briefs, style notes, and composition requirements.

Built by OpenAI

the team behind ChatGPT and Sora — and ready to drop into automated creative workflows.

Best for

Marketing campaign visualsProduct concept mockupsSocial media creativesAd variations and A/B testingEditorial and blog imageryReference-guided image generation

How GPT Image 2 works

GPT Image 2 takes a text prompt and, when provided, reference images to generate a new image that follows the requested subject, style, composition, and format. Users choose resolution and aspect ratio settings, then submit the prompt and optional image inputs for generation. The model returns image outputs suitable for creative production, iteration, and marketing workflows.