Using GPT-4o To Make Stunning, Original Illustrations in ANY Style

In partnership with

Find out why 1M+ professionals read Superhuman AI daily.

In 2 years you will be working for AI

Or an AI will be working for you

Here's how you can future-proof yourself:

1. Join the Superhuman AI newsletter – read by 1M+ people at top companies

2. Master AI tools, tutorials, and news in just 3 minutes a day

3. Become 10X more productive using AI

Join 1,000,000+ pros at companies like Google, Meta, and Amazon that are using AI to get ahead.

Sign up and start learning AI

Making Original Illustrations in ANY Style

Check out this photo I snapped on my cellphone of some tofu (no, this isn’t the setup for an April Fools joke—I HATE April fools day and will not subject you to that).

It’s fine. You probably have a thousand just like it on your own phone. If I was writing a food story and needed an illustration, it would work okay.

Now, take a look at this bespoke, minimalist watercolor painting.

Using illustrations like that would really make my article stand out.

And with GPT-4o Image Generation, it’s now super easy to go from the first one (a basic and fairly generic cellphone photo) to the second (a beautiful, original illustration in any visual style).

Want my tofu as a gritty, Film Noir image? Boom, here it is!

The Power of Transformations

If you’ve seen the viral Studio Ghibli memes circulating online, you know that the new GPT-4o Image Generation is fantastic at transforming raw images into new visual styles.

Making your cat into a cartoon character is fun. But it won’t pay the bills.

• Illustrating your Medium articles with bespoke, original watercolor images that make them stand out—and drive clicks?

• Updating old blog posts by turning tired and outdated imagery into brand new, original illustrations?

• Taking a public domain image or something you found on Unsplash and making it into an original line drawing for use in your next video?

Those things genuinely will. And GPT-4o makes transforming images super easy.

A quite note on why this is new. Before, you could hand GPT-4o a style reference image, and then ask for a new illustration in the same visual style.

But it was always a generic illustration—not something specific.

You could ask for a Millennial age AI expert in San Francisco with blond hair and a penchant for wearing sunglasses indoors. But you couldn’t ask for “Thomas Smith.”

Now, you can just hand GPT-4o an image of me, and ask it to transform it into something new. The move from general illustrations to specific illustrations is the secret sauce that makes GPT-4o Image Generation amazing for creators.

Here’s a video I made that walks through my exact process for using it to transform images:

For creators, this kind of specificity is huge. Readers love original imagery. Algorithms love it even more.

Now, you can call up an unlimited number of highly specific, original illustrations that fit the mood of whatever you’re creating—for free.

Sure, make a Ghibli cat for your own consumption. But for your content, try these techniques to add something truly different—perhaps even beautiful—to whatever you’re making.