OpenAI launches new Image generation model 'GPT-Image-1' under API for Developers

On April 23, 2025, OpenAI unveiled its latest image generation model, now available via API under GPT-Image-1. According to the AI giant, the model marks a major leap in AI image generation capabilities. The model is the same one powering ChatGPT’s image functionality and boasts high versatility. The model can create images in different styles, work on custom instructions, use references from the internet, and render texts in the images correctly.

Also read : Open AI’s Sora is just too good that it might’ve been trained in Unreal 5

The new GPT-Image-1 model opens up a broad spectrum of applications across industries like design, advertising, education, and entertainment. Pricing is tiered based on usage: $5 per 1 million text tokens, $10 for image input tokens, and $40 for image output tokens. That means, depending on image quality and size, each generated image could cost as little as $0.02.

To help developers get started, OpenAI has released a refreshed image guide, a sample gallery, and updated documentation including SDKs with options for adjusting quality, format, size, and moderation levels. Notably, access to the model requires organizational verification to ensure responsible use.

More integrations are on the way, including support via the Responses API. With the rapid rollout, it’s clear OpenAI is setting the stage for an even more visual AI future.