We tested 5 AI tools to help you make money with your crafts.

Four gave us product ideas worth exploring. The fifth gave us a cat and some questions.

You guys know we love making cool stuff. But if you bought a machine for your business, eventually you have to ask the question that matters: What can I make that somebody will actually pay for?

So we opened the AI design tools inside Atomm, fired up the xTool O1 Omni UV printer and xTool P3 laser, and made five very different projects. We went from a regular photo of Garrett standing in the hallway to a custom baseball card, turned our family into a Christmas keepsake, and gave a challenge coin some seriously impressive texture.

We also made a shadow box that may need a little more time to find itself. Let's talk about all of it.

Watch the full test

See the designs, the machine setup, the finished projects, and our honest reactions in the video below.

Watch the video on YouTube

How we tested the five tools

For each project, we asked the same three questions:

1. How easy was it to create the design?
2. How easy was it to produce the finished item?
3. Would we sell it in our own business?

Our difficulty scale runs from 1 for easiest to 5 for hardest. We were testing the path from idea to a physical product and deciding where we saw selling potential. We weren't running a sales experiment or calculating profit margins.

The O1 Omni handled the UV printing. The P3 cut the acrylic cards and ornament after printing, and it cut the wooden shadow box. That last project was a laser-cut test; we didn't UV print the wood.

1. Sticker Maker: small products with plenty of personality

Stickers were our easiest place to start. We chose the cartoon option in Atomm's Sticker Maker, uploaded a photo of the two of us, and let it create a few variations.

The result? Tiny cartoon Kim and Garrett, being considerably cuter than two people who have been standing around a printer all day.

How we made them

We opened the designs in xTool Studio, selected the Omni, resized the artwork, and chose the regular transfer sticker material for our AB film. The imported file included both a bitmap image and a vector layer. We turned off the vector output for this print job, then printed, manually laminated the sheet, and separated the stickers by hand.

That manual laminating step was the reason production scored a little higher than design. The artwork was easy; finishing the sheet added another task.

Where we'd look for customers

Think about sticker packs for a particular group: dance parents, a local business, a team, or customers who love a very specific kind of humor. Stickers can also be an add-on to a larger order or a branded freebie that makes a package more memorable.

Our favorite business question here is, “Who would see this and immediately say, ‘That's so me’?” That gives you a much clearer starting point than making a pile of unrelated designs and hoping somebody adopts them.

Our verdict: Design 1/5. Production 2/5. Would we sell them? Absolutely. We already sell stickers, and this was a useful way to create more personalized artwork.

2. Commemorative Coin Generator: the texture got our attention

We already make UV printed challenge coins, so this tool had a little more to prove. We selected a constellation style in the Commemorative Coin Generator and asked for a Virgo design.

The finished coin had stars, a moon, a detailed central figure, and raised relief that made us want to keep tilting it in the light. Even the moon had dimension. This was one of those projects where the close-up really matters.

The setup tip we'd use next time

Download the generated artwork, place the blank on the printer bed, and run Smart Mask before importing the image. In our test, importing the artwork first meant we had to resize and position it manually. Scanning first would have let the imported artwork snap to the detected coin.

We also removed the protective film from the blank and selected the brass commemorative coin material. The imported coin design brought in a relief texture, which gave us that raised finish.

Who might buy a custom coin?

Local businesses, car clubs, first responders, military groups, sports teams, conferences, and special events all came up in our discussion. The appeal is creating a coin for a particular group or occasion, especially when they want a smaller custom run.

A club anniversary or a meaningful team milestone gives the design a reason to exist. That's a stronger offer than “we have a machine that prints on circles.”

Our verdict: Design 1/5. Production 1/5. Would we sell them? Yes. The relief detail made this a standout result.

3. Card Generator: Garrett gets his baseball comeback

This project started with Kim asking Garrett to stand in the hallway for a picture. No stadium. No uniform. No dramatic sports photography. Just Garrett, a phone, and the hope that AI could handle the rest.

We uploaded the photo, entered his name, number 9, position shortstop, and team Bulldogs. The Card Generator turned that everyday picture into a baseball card with an action pose and a whole lot more athletic ambition.

Printing on one machine, cutting on the other

We used the multi-device print-and-cut workflow in xTool Studio. Here's the sequence we demonstrated:

1. Import the card artwork and resize it to about 3.5 inches tall.
2. Create a vector outline with a zero offset for the card's cut line.
3. Assign the bitmap artwork to the O1 Omni and the vector cut line to the P3.
4. Scan the printer bed, position the artwork on the white acrylic, and print with the alignment marks.
5. Move the acrylic to the P3, use Auto Alignment to locate those marks, and cut the card.

Watching the cut outline move into place was one of our favorite parts. There are two machines involved, but the alignment step made the handoff feel straightforward.

And then we made ourselves a couples card

Of course we tried the couples option. Ours included our anniversary date and a little “Balance & Chaos” energy, which seems appropriate for two people who work together and willingly keep buying equipment.

We added a magnet to the back of the couples card. That gave the same design another use: a personalized keepsake for the refrigerator.

For a business, we'd explore a team package or a partnership with a local card shop. Parents and grandparents could order personalized cards of their player. The same idea could work for cheer, dance, martial arts, pets, or a wedding party.

Our verdict: Design 1/5. Production 1.5/5. Would we sell them ourselves? Probably not, because cards aren't our niche. But we could absolutely picture this fitting a business with the right customers.

4. Fridge Magnet Generator: surprise, it's also an ornament

The tool said fridge magnet. We saw a Christmas ornament. Then we added a magnet anyway, because apparently this project wanted options.

We started with a cartoon image of our family, selected a holiday theme, added “Merry Christmas,” and used an ornament shape. The resulting design brought the family together in matching Christmas sweaters with a bow and holly at the top.

We made this sample about four inches tall on white acrylic, added a small offset around the artwork, and created a hanging hole. Then we used the same Omni-to-P3 print-and-cut process as the cards. We chose the texture option for some dimension, though the effect was less raised than the relief on the coin.

The useful lesson here is to look at what a generator creates and imagine where that design could go. A family image could become an ornament, a gift tag, a bag tag, a locker decoration, or a keepsake magnet.

For customers, we'd start with personalized family gifts: a grandparent ornament featuring the grandkids, a pet keepsake, or a holiday design made around one family's photo. The personal connection gives someone a reason to choose yours.

Our verdict: Design 1/5. Production 1.5/5. Would we sell it? Yes. This was one of the clearest gift ideas in the test.

5. Layered Cutting Generator: the cat was cute

Our final test was a Halloween shadow box. We asked for a black cat on a fence in a graveyard, with a full moon, and selected four layers.

That sounds like something we should love. After roughly six or seven attempts, though, we still weren't getting a composition we were excited about.

The generator did create an SVG that was easy to import and cut on the P3. Production wasn't the problem. The design had awkward layers, a lot of visual weight near the bottom, and open space that didn't feel intentional.

We could imagine adding words or rebuilding parts of the scene, but that would require more design work. For this particular test, the AI hadn't gotten us to a finished product we'd be happy to put on the table.

Our verdict: Design about 4–5/5 after repeated attempts. Production 1/5. Would we sell this version? No. The tool produced a cuttable file, but we still needed to judge whether the finished design was good enough.

Our five-tool scorecard

Here's how the projects compared in our shop. Lower difficulty scores mean easier. On a small screen, scroll across the table to see every column.

Turning a neat project into something someone buys

The encouraging part of this test was how quickly several tools helped us move from an idea to usable artwork. That first design step can be a real sticking point when you're learning a machine and trying to build a product line at the same time.

But you still need to decide who the product is for. A dance mom sticker, a Little League card for proud grandparents, a family ornament, or a commemorative coin for a local club gives you a specific person to talk to and a specific reason they might care.

If we were choosing one of these ideas to test with customers, here's where we'd start:

1. Choose one customer group. Start with a team, club, local business, or community you can actually reach.
2. Make a sample that fits that group. Use the colors, names, occasion, or inside joke that makes the product feel personal.
3. Describe a clear offer. A personalized player card with a matching magnet is easier to picture than a long list of everything your printer can do.
4. Work out the full cost. Include the blank, ink, finishing supplies, packaging, AI credits, design revisions, machine time, and your hands-on work. Atomm image generation used credits in our test, so those belong in the calculation too.
5. Get feedback before making a mountain of inventory. Show the sample to likely buyers, learn what they'd change, and test a small offer before committing to a big run.

A tool that saves you design time can be valuable. A product your customers actually want is what gives that saved time somewhere useful to go.

Which one would you try first?

Would you start with stickers, textured coins, custom cards, or a personalized ornament? Or do you have a shadow box prompt that could redeem our poor cat?

Tell us in the comments on the video which product you think would sell and who you'd make it for. You guys might give us the idea for part two.

Watch the video & tell us your pick