OpenAI just gave its image generator a brain upgrade. Images 2.0 — dropping exactly a year after the original — doesn’t just generate pictures anymore. It researches, synthesizes web information, and turns all of that into what OpenAI calls “production-level” visuals. So instead of prompting it to draw something you already have in your head, you can apparently point it at a topic and let it figure out what the visual should even look like.
The editing side got a serious push too. Background removal, aspect ratio adjustments, up to eight images at once in 2K resolution. OpenAI’s clearly going after the people who currently bounce between Photoshop, Canva, and whatever else just to finish one project.
Infographics and diagrams are supposedly much sharper now — better multilingual text handling, tighter prompt adherence, and fewer of those annoying moments where it ignores half of what you asked for.
One split worth knowing: the basic Images 2.0 is available to all ChatGPT users, but the “thinking” mode — the part that actually reasons through the image before generating it — is locked to Plus, Pro, Business, and Enterprise plans.
