OpenAI Previews ChatGPT Images 2 Generation Model with Real-World Intelligence

OpenAI has shown the world a new generative model called ChatGPT Images 2, and it is more than just your average image-generating model.

The move comes right after OpenAI announced the shutdown of Sora, the company's generative AI video model.

OpenAI Previews ChatGPT Image 2: New Image Gen Model

OpenAI revealed in its latest livestream event that there is an upcoming upgrade to its existing generative model with the new ChatGPT Image 2, which has improved capabilities in creating content.

Previously, OpenAI only gave users the ability to generate images via the ChatGPT Image model, but now, based on its teaser before the event and during the showcase itself, it can create magazine designs.

Through this new feature, users may basically create a magazine page or even an entire magazine via the Image 2 model.

Additionally, the ChatGPT Image 2 model can also deliver images featuring 2K resolution in multiple aspect ratios, generate images using updated data from the web (up to December 2025), feature Instant and Thinking modes, and add non-Latin languages to its expertise, according to 9to5Mac.

When to Expect the New ChatGPT Image 2 Model?

OpenAI is rolling out the latest ChatGPT Image 2 model now, and users may try it once they see it available on the app or website platform.

As previously mentioned, OpenAI gave ChatGPT Image 2 real-world intelligence because of its updated knowledge cutoff of up to December 2025, which is combined with handling end-to-end tasks.

Initially, OpenAI's solution to ChatGPT's image generation capabilities was sharing DALL-E's generative models on the platform, but the company later changed it to the ChatGPT Image model in December 2025.

It was revealed to be the company's solution to the massive popularity of Google's Nano Banana Pro.