OpenAI just set the AI world on fire again—this time by rolling out a brand-new image generation capability within GPT‑4o that has users everywhere buzzing.
This isn’t your ordinary AI image generator, either. Built directly into the GPT‑4o model, it’s opening up a radically new era for how we create, edit, and refine images in ChatGPT.
we are launching a new thing today—images in chatgpt!
— Sam Altman (@sama) March 25, 2025
two things to say about it:
1. it's an incredible technology/product. i remember seeing some of the first images come out of this model and having a hard time they were really made by AI. we think people will love it, and we…
How do you get access? And how is this going to change design as we know it?
I got the scoop from Marketing AI Institute founder and CEO Paul Roetzer on Episode 142 of The Artificial Intelligence Show. And, based on his hands-on experience, this new image generator is making previous AI art tools look like child’s play.
Here’s everything you need to know.
Why GPT‑4o Image Generation Is a Big Deal
First off, this new image generator is seamlessly integrated into GPT‑4o. As a result, it goes well beyond the older DALL·E-type tools we’ve all tried in the past. According to OpenAI’s launch announcement, you can now precisely render text in your images, add or remove elements in existing photos, and refine your visual output through a natural conversation with ChatGPT.
“It is definitely quite impressive,” says Roetzer.
Why? Because GPT‑4o is natively multimodal. That means the model’s full intelligence is brought to bear on your prompts, giving you more accurate and more flexible results. It’s better at handling text on images, too—a longtime Achilles’ heel for older models.
Early testers (including Paul) say the results are indeed stunning, with the new image generator nailing complex text in images and consistency across images that stymied previous models.
In other words, you can effectively talk your way to a final, polished image—and keep refining it with each conversation turn—without entirely losing consistency or style from one version to the next.
What This Means for Creatives, Brands, and Businesses
If you’ve ever spent days (or weeks) going back and forth with designers on a simple creative concept, GPT‑4o’s new capabilities might feel like magic. You can now produce highly detailed, iterative mock-ups of logos, ads, or entire brand assets—on your own.
That doesn’t mean professional designers instantly vanish. But it does mean you can get to first (or second, or tenth) draft way faster, then bring in the experts for finishing touches.
“You're going to have the ability to do the first drafts yourself now for anything,” says Roetzer. “And you still may rely on the experts to do the final products and bring it home, but some of that early work might just be done by the AI.”
On the flip side, businesses may start raising their expectations for how quickly and cost-effectively creative work can get done. After all, if a single marketing manager can spin up dozens of on-brand ad variations in mere hours, why wait for days or weeks?
Roetzer says it becomes “quite apparent” the moment you use these tools that they’re going to have a significant impact on creative work. But what that means long-term for these professions is less clear.
“All of a sudden non-designers have these abilities and I don’t know what that means, honestly,” he says. “I don’t think OpenAI knows what it means. I don’t think Google knows what it means. But I think it’s really important that we have these conversations, because I just feel like these tools are starting to truly creep in to democratize the ability to build things.”
Video Could Be Next
As jaw-dropping as GPT‑4o’s new image skills are, they may just be a warm-up for something even bigger: true AI-driven video generation.
OpenAI hasn’t announced anything official yet in that department, but Paul has some predictions:
“Imagine this level of control and consistency, but applied to 10, 15, 20 second videos,” he says. “I have to imagine when the GPU shortage sort of goes away and they have more capacity, that capability's probably already sitting in there. They just don't have enough GPUs to roll it out.”
We’ve already seen video-generation releases from players like Google (with its own advanced research on generative video). As those tools get more robust—and OpenAI leaps in with an offering of its own—there’s a good chance you’ll have a fully integrated text, image, and video creation suite inside ChatGPT.
Don’t Have Access Yet? You’re Not Alone…
The new image generation feature is currently only available to ChatGPT Plus, Pro, and Team users. That means it might be a bit before free-tier users get a chance to try it out. Sam Altman even mentioned that OpenAI’s GPUs are “melting” under the massive influx of usage—so the expansion to all users could take some time.
When you do finally get your hands on it, expect to find the interface under the same ChatGPT environment. You simply describe what you want, refine with follow-up prompts, and watch GPT‑4o handle the rest.
The Bottom Line
GPT‑4o image generation is one of the most powerful signals yet that AI isn’t just about words anymore. It’s about seamlessly fusing language and visuals into a single creative workflow, which could forever change how we conceptualize, design, and iterate on digital or physical products.
In Paul’s view, we’re witnessing “first draft” AI capabilities, but they’re already surprisingly strong. And that begs a larger question: When the tool can produce consistent, refined results that blend text, imagery, and soon (maybe) video, how will that reshape the roles of creative teams—and the future of work itself?
No one has all the answers to that yet. But if you spend a few minutes in GPT‑4o’s new image generator, you’ll get a taste of just how drastically things may change—faster than most organizations are prepared for.
“These capabilities are significant and you can definitely start to imagine a world where you’re using AI more and more in creative work.”
So buckle up, because image generation is only the beginning. AI-fueled creativity just went into overdrive—and there’s no turning back.
Mike Kaput
As Chief Content Officer, Mike Kaput uses content marketing, marketing strategy, and marketing technology to grow and scale traffic, leads, and revenue for Marketing AI Institute. Mike is the co-author of Marketing Artificial Intelligence: AI, Marketing and the Future of Business (Matt Holt Books, 2022). See Mike's full bio.