1 What Is Graphic Design
1.1 What you can make
Graphic Design is Atomm AI's AI design workspace built for makers. On a single canvas you can generate, convert, edit, and prepare for production — and the output is made for the way you'll actually fabricate it:
| Output | Typical products |
|---|---|
| Engraving artwork | Portrait and pet-photo engravings on wood, slate, leather, metal, or tumblers; logos; line art; lettering on cutting boards |
| Cut files | Cake toppers, door signs, openwork decor, silhouettes, connected lettering, layered wall art, bookmarks |
| Print artwork | T-shirt prints, UV prints, pet portraits, seamless patterns, greeting-card covers |
| Print-then-cut files | Sticker sheets, die-cut fridge magnets, acrylic keychains and pendants, coasters, business cards, badges |
| Vector | Photo-to-line-art, color illustrations with editable color blocks |
| Relief | Grayscale depth maps for coins, badges, and photo reliefs; color print reliefs with gem, knit, embroidery, or fossil textures |
| 3D | 3D Photos or 3D Models of people, pets, landmarks, vehicles, and flowers for crystal or clear-acrylic engraving |
1.2 How it differs from a generic AI image generator
What sets Atomm AI apart from a generic AI image tool is the built-in craft know-how. It understands not only what picture you want but also how you're going to make it: which structures cut cleanly, which images engrave well, when a vector is needed, and what kind of output each process requires.
So you don't have to master prompt engineering, image-editing software, or fabrication rules first.
Describe it, we'll make it craftable.
1.3 Graphic Design or Structure Build?
Atomm AI has two design modes. This guide covers Graphic Design. If you want to build a 3D object that is assembled from parts, see the Structure Build guide.
| Use Graphic Design | Use Structure Build (see its own guide) |
|---|---|
| Make an image to engrave, cut, print, or engrave inside crystal | Make an object that is laser-cut from panels and assembled |
| Door signs, stickers, T-shirt prints, coins, crystal portraits | Storage boxes, drawers, shelves, stools, lampshades |
| Output is a bitmap, vector, depth map, or 3D model | Output is a 3D structure + 2D cut layout + adjustable parameters |
| You want to modify an existing photo, logo, or design | You want a brand-new structure generated from a description |
 |  |
1.4 The workspace at a glance
When you open Graphic Design you'll see five areas. Every feature in this guide lives in one of them.
| # | Area | What it's for |
|---|---|---|
| ① | Left toolbar | Ways to add content to the canvas: Upload, AI chat, Template, Text, Elements, AI Font, Mockup; Layers at the bottom |
| ② | AI chat panel | Where you talk to the Agent. History and Memory at the top; "Explore these design ideas" recipe categories in the middle; the input box at the bottom, where you can attach images, open the recipe library, switch Fast / Quality, and set ratio and count |
| ③ | Canvas | Everything you generate or upload appears here. Select an object and its quick tools (such as BG Remove, Vectorize, Laser Prep) appear above it |
| ④ | Settings panel (right) | Properties of the selected object: size, alignment, Color Adjustment, Image Adjustment, Fill / Stroke, and more. Shows "Select an element to edit" when nothing is selected |
| ⑤ | Export | Top right. Exporting files, opening in xTool Studio, and publishing to the Atomm community all start here |
1.5 A simple mental model
Graphic Design has three kinds of capability, each with one job:
- The Agent (chat) understands and moves the work forward. You say what you want; it generates, converts, and revises, and it remembers what has happened in this conversation.
- Recipes get you started fast. Each one is a pre-tuned plan for a specific product; you only fill in the required photo and text.
- Canvas tools refine. Once the design has direction, use them to adjust one specific property precisely.
These aren't three parallel paths — they're the tools you reach for at different stages of the same piece. You usually start with chat or a recipe and finish on the canvas.