OpenAI

GPT Image 2

Name: OpenAI GPT Image 2
Author: OpenAI

Compare

2026-04-21

Compare

GPT Image 2 is OpenAI's next-generation image model, the successor to GPT Image 1.5. Built on an entirely new architecture independent from GPT-4o, it topped the LM Arena image leaderboard at launch. Text-rendering accuracy reaches 99%+, with high-fidelity glyph support for Japanese, Korean, Chinese, Hindi, and Bengali, enabling the creation of clean infographics, slides, posters, maps, UI mockups, and manga. Outputs reach up to 4K at custom dimensions, with three quality tiers and two operational modes: Instant for fast generation and Thinking for reasoning- and web-search-integrated tasks.

Vision|Proprietary Model

Knowledge Cutoff

Unknown

The date this AI finished learning. It may not know about things that happened after this date.

Input → Output Format

The types of content this AI can receive, and what it can produce in return.

Context Memory

—

The maximum amount of text the AI can read and process in a single request. A larger number means it can handle longer documents or conversations.

Cost/1M Words

$8IN$30OUT

The cost of using this AI directly in your own application. Shown in USD per 1 million units of text (tokens).

Calculate Cost

Source:Official Docs

AI Performance Evaluation

Arena Image Score

1513

±8

As of 2026-04-21

Overall Rank

🥇 No.1

15,127 Votes

Arena by Image Category

Art

1422±23🥇 No.1

Cartoon

1543±12🥇 No.1

Photorealistic

1493±12🥇 No.1

Portraits

1554±18🥇 No.1

Text Rendering

1574±14🥇 No.1

3D Modeling

1490±24🥇 No.1

Commercial Design

1523±14🥇 No.1

Source:Arena Intelligence

OpenAI