You finished the ebook. The writing’s done, the PDF’s exported, the sales page is half-built.
Then you hit the part that virtually every new digital product needs (or at least deserves): you need it to look like a real product.
So you go looking. The decent mockup tools want $29 a month. The free ones hand you something that looks like a 2009 clip-art disaster with a drop shadow bolted on.
You don’t need either one. You need two prompts and about four minutes.
I ran this on my own product this morning — The Art of Caching — and I’m printing both prompts below, exactly as I used them. Copy them. They work.
What You Walk Away With
A photo-real 3D book mockup of any ebook you own. No Photoshop. No subscription, and honestly very little effort.

Why Two Prompts Instead of One?
Fair question. Why not just ask for the finished mockup in one shot and skip a step?
Because I tried that. And it broke in a way you would potentially miss without noticing.
I asked for the whole thing at once — 3D book, desk, lighting, title, subtitle, all of it in a single prompt. Got back a gorgeous image. Genuinely good lighting, believable spine, nice shallow depth of field.
Subtitle read: “A PREPPER’S GUIDE TO TO HIDDEN READINESS.”
Two “TO”s. I ran that image past three different vision models to make sure I wasn’t imagining it. All three read it back the same way. That’s not to say it will always occur, but there is a higher chance that it will.

Here’s the mechanics of it. When a model renders text and geometry at the same time, the text is competing for attention with perspective, curvature, shadow, and depth. Something loses. It’s usually the text.
Give it a flat cover to work from and the text isn’t being generated anymore. It’s being moved. Totally different job, and models are much better at it.
That’s the whole trick. Separate the typography from the geometry.
Step 1: Build the Flat Cover
Paste this into an image generator that handles text well. Swap the bracketed parts for yours.
The last line is doing more work than it looks like. “No mockup, no book spine, no shadow.”
Leave that off and the model helpfully gives you a 3D book — which is the exact thing you’re trying to avoid at this stage. You want flat art. Boring, straight-on, edge-to-edge flat art. Prompt two needs a clean canvas to give us the best results.
Set the aspect ratio to 2:3. That’s real book proportions. Square covers scream “I made this in an AI tool.” (although really nowadays I assume most are AI created).
Step 2: Wrap It Into a 3D Mockup
Now take that flat image and feed it in as the source. Paste this:
That’s it. That’s the product shot.
The Honest Part
I put that “keep the title text exactly as provided” line in there expecting it to be the hero of this article. The magic clause. The thing I’d tell you never to skip.
So I tested it without the line. Same result. Text came back perfect.
What that tells me is the clause isn’t what likely saves you — the flat cover is. Once the typography already exists as pixels, the model has nothing to reinvent. The instruction is cheap insurance, so keep it. Just don’t believe it’s doing the heavy lifting. The More You Know.
I’d rather tell you that than sell you a magic word that isn’t magic.
Which Tools, and Why Each One
Step 1 — anything strong at text. I used Ideogram (though I have good results with Nano and even some others). It’s the most reliable thing going for clean lettering, and a cover lives or dies on its title. ChatGPT’s image tool works too. Whatever you use, the requirement is the same: it has to spell.
Step 2 — an image editing model, not a generating one. I used Nano Banana, which is Google’s image model inside Gemini. This distinction matters more than the brand name. You need something that takes your image and transforms it. Hand this job to a pure text-to-image generator and it’ll cheerfully invent a whole new cover and hand it back like it did you a favor.
Free tier on the Gemini app covers this. That’s the $0 in the title.
Now Milk It
One flat cover feeds every product image you’ll ever need. Keep prompt two, change the last sentence:
Tablet version. “Display this cover on a tablet screen held in one hand, coffee shop background, blurred.” Instant digital-product shot.
Bundle stack. “Three copies of this book stacked at slight angles on a marble surface, top copy face-up.” Makes one ebook look like a collection.
Device spread. “This cover shown simultaneously on a laptop, tablet, and phone arranged on a wooden desk.” The classic PLR sales-page hero.
Same source art every time, so your product looks consistent across the sales page, the emails, and the thumbnails. Which is what actually makes a product look legitimate — not the shadow quality.
Go Do It
Pick an ebook you already own. A PLR file, an old lead magnet, something rotting in a folder you long ago abandoned.
Run prompt one. Run prompt two. You’ve got a product shot in four minutes.
Then go cancel the mockup subscription.
