Unifically LogoUnificAlly
Model logo

GPT Image 2 API

Text to ImageImage to Image

Latest OpenAI image generation with reference image support. Medium and high quality options.

Click or drag & dropPNG, JPG, WEBP, GIF · Max 100MB
Output

Your generated image will appear here

Features

What GPT Image 2 API offers

Same integration surface as GPT Image 1 and 1.5: prompt, aspect ratio, quality, and optional reference images
Reference-guided generation with up to 16 images per request
Aspect ratios: 1:1, 2:3, and 3:2
Quality: medium or high (no low tier)
Latest GPT Image generation tuned for the cleanest detail and tightest prompt adherence in the family
Async JSON API with separate generate and status routes
Usage-based billing on Unifically

Use cases

Built for

Primary

Brand and campaign hero images - Final-quality stills when you want the best fidelity OpenAI offers

#2

Packaging and retail mocks - Legible type and controlled composition with up to 16 reference uploads

#3

Product marketing - Cleaner reflections, materials, and subject isolation than older GPT Image tiers

#4

Pitch decks and one-pagers - Text-forward visuals without spinning up a full design pipeline

#5

Iterative creative - Same parameter shape as GPT Image 1 / 1.5 so swapping models requires no code change

#6

Asset libraries - Batch prompts where consistency matters more than the lowest possible price

FAQ

About GPT Image 2 API

GPT Image 2 is the newest tier in OpenAI's GPT Image family on Unifically. It accepts the same request fields as GPT Image 1 and 1.5: prompt, optional aspect_ratio, optional quality, and optional image_urls for references.

GPT Image 2 is the newer model aimed at the cleanest detail and tightest prompt adherence in the family. Both 1.5 and 2 expose the same controls in the playground, so you can A/B test outputs without changing your integration.

GPT Image 2 ships with medium and high quality only. The low tier was dropped to focus on the model's production-grade output.

You can include up to 16 reference images in one call, each up to 100 MB, matching the GPT Image 1 and 1.5 limits.

Requests are billed through Unifically using the openai/gpt-image-2 price key. Check the model pricing table for the current rate.