← Models

GPT Image 2

OpenAI · GPT Image

OpenAI's current state-of-the-art GPT Image model behind ChatGPT Images 2.0 and API image generation.

Part of GPT Image family · Other versions: gpt-image-1-mini , GPT Image 1.5 , GPT-Image-1
Type
image
Context
N/A
Max Output
N/A
Status
current
API Access
Yes
License
proprietary
image-generation image-editing chatgpt-images multilingual-text visual-reasoning api openai
Released April 2026 · Updated April 24, 2026

Overview

Freshness note: Image-model capabilities, safety behavior, and pricing can change quickly. This profile is a point-in-time snapshot last verified on April 24, 2026.

GPT Image 2 is OpenAI’s current state-of-the-art image generation model, launched with ChatGPT Images 2.0 on April 21, 2026 and listed on OpenAI’s API pricing page as gpt-image-2. It succeeds GPT Image 1.5 as the most important GPT Image family entry for new OpenAI image work.

The product framing is broader than “make a picture.” ChatGPT Images 2.0 is aimed at usable visual artifacts: posters, infographics, multilingual layouts, manga pages, product mockups, classroom visuals, campaign assets, and other outputs where text rendering, instruction following, and composition matter.

Capabilities

GPT Image 2 improves the practical parts of image generation that often break real workflows: dense text, multilingual scripts, small labels, UI-like layouts, visual consistency, aspect-ratio flexibility, and style control. OpenAI also presents thinking mode as a way for ChatGPT to reason, search, and prepare before generating visuals.

That matters for teams building visual workflows because the bottleneck is rarely one pretty image. It is getting images that preserve details, follow a design brief, produce readable text, and can be iterated without restarting from scratch.

Technical Details

This is an image-native model, so contextWindow: 0 and maxOutput: 0 are intentional in this repository. They should be read as N/A for language-model token comparisons, not as literal capability limits.

OpenAI’s pricing page lists separate token rates for image inputs, cached image inputs, image outputs, text inputs, and cached text inputs. The ChatGPT Images 2.0 system card also describes thinking mode as adding reasoning and tool use to the image generation process.

Pricing & Access

OpenAI’s pricing page lists gpt-image-2 as available under multimodal models with:

  • Image input: $8.00 per 1M tokens
  • Cached image input: $2.00 per 1M tokens
  • Image output: $30.00 per 1M tokens
  • Text input: $5.00 per 1M tokens
  • Cached text input: $1.25 per 1M tokens

Actual per-image cost depends on size, quality, input images, edits, and whether the workflow uses text or image tokens. Do not compare it directly to LLM text pricing without modeling the generated image-token count.

Best Use Cases

Use GPT Image 2 for professional image generation where detail, layout, readable text, and iteration quality matter: ads, posters, presentation visuals, infographics, product mockups, branded concept directions, comics, and educational diagrams.

For low-stakes bulk variation, cheaper or older image routes may still be enough. For final production design, human review remains necessary.

Comparisons

  • GPT Image 1.5 (OpenAI): Previous flagship GPT Image tier, still useful as compatibility context.
  • GPT Image 1 mini (OpenAI): Better fit for lower-cost, higher-volume generation.
  • Imagen 4 (Google): Strong API-backed image alternative, with ecosystem and style tradeoffs.
  • Nano Banana 2 (Google): Relevant comparison for Gemini-native image workflows and consumer-facing creative use.