Paste this into your AI tool
Generate this design to professional print specifications: • Final size: decide the printed size first, then use a tool/plan that can render it — 24 in wide × 36 in tall. • Bleed: add 1/8 in (0.125 in) on every side. Extend the background and any edge artwork into the bleed — no white border. Total canvas: 24.25 in × 36.25 in. • Resolution: generate at the HIGHEST resolution your tool and plan allow — up to 600 DPI (14,550 × 21,750 px) when the tool supports it. More pixels is always better, never a problem. Never go below 3,638 × 5,438 px (~150 DPI), the minimum for this size — big prints are viewed from farther, so ~300 DPI suits something held in hand, ~150 a poster, ~100 a large banner, but always aim higher when the tool allows. If it can't reach the size, make the most it can, then upscale it (see "Going bigger" below) — or we'll enlarge it. • Safe zone: keep all text, logos and anything important at least 1/4 in (0.25 in) inside the trim edge. • Color: use bold, high-contrast color. Avoid bright neon greens, blues and oranges — they shift in CMYK. Deep blacks solid, not gray. • Text: don't bake important words (names, prices, phone numbers) into the image — AI often misspells them. Add that text yourself, or we'll set it cleanly. • Detail: no hairline-thin lines; no text below ~6 pt. • File: highest-quality lossless PNG or TIFF — or vector SVG/PDF for flat art. Avoid heavy JPEG compression. Note: for very large posters or banners, pick a tool built for size — vector (Recraft) scales infinitely, Stable Diffusion tiles to huge pixels, or use a hi-res / upscale mode. AI outputs RGB and can't add crop marks or true CMYK — that's expected; PrintSimply converts to CMYK and adds bleed + crop marks for you.
Your size → the numbers
Enter the final printed size and how close it's viewed. We fill the prompt with the bleed canvas and pixel target.
Always output the most pixels your tool allows — more is never a problem.
See how the print specs — full-bleed, the 24 × 36 in trim, the 0.125 in bleed, the resulting 24.25 × 36.25 in document, and 300 DPI — are written right into the prompt. Swap in your own subject and size. After you generate the image, save it as a PDF at 24.25 × 36.25 in (drop it into a page that size in Canva or Word and Export as PDF) — or send us the image and we'll make the print-ready PDF.
The same example, written for each AI tool
New to AI? ChatGPT and Canva below are the easiest — just follow the numbered steps. Whichever tool you use, copy the prompt and send us the result; we'll do the rest.
- On a computer or phone, open chatgpt.com (or the ChatGPT app) and sign in — a free account works.
- Press the Copy button above to copy the prompt.
- Tap the message box at the bottom, paste, and press send.
- Wait a moment for your picture to appear.
- Save it — on a computer, right-click the picture → Save image; on a phone, press and hold it → Save.
- Open canva.com and sign in — a free account works.
- Click Create a design → Custom size, type 24 in × 36 in, and click Create.
- In the top File menu, switch on Show print bleed.
- On the left, click Apps → Magic Media. Press Copy above, paste the prompt into the box, and click Generate.
- Drag a picture out until it covers the whole page, past the dotted lines.
- Click Share → Download, choose PDF Print, tick Crop marks and bleed, and download.
- On a computer, open aistudio.google.com and sign in with a Google account (free).
- At the top, choose the model Nano Banana Pro (or Nano Banana 2).
- Set the shape to 2:3 (tall) and the size to 4K (the biggest).
- Press Copy above, paste the prompt in the box, and click Run.
- When the picture appears, click it and download it.
- Open ideogram.ai and sign in (free).
- Set the shape to 2:3 (tall) and pick the highest quality.
- Press Copy above, paste the prompt, and click Generate.
- When the picture appears, click Upscale to make it bigger, then download it.
- Open recraft.ai and sign in (free).
- Choose the Raster style for a photo look (or Vector for flat, cartoon-style art).
- Set the canvas shape to 2:3 (tall).
- Press Copy above, paste the prompt, and click Generate.
- Click Download and choose PDF.
- This one is for people already using Midjourney. If it feels tricky, use ChatGPT or Canva above instead.
- Press Copy above and paste the whole prompt (the --ar 2:3 at the end sets the tall poster shape).
- Send it, then click Upscale on the picture you like.
- Save the upscaled picture.
- This is a tool for advanced users. If that's not you, use ChatGPT or Canva above.
- Set the picture shape to 2:3 at the biggest size your computer allows (for example 1024 × 1536).
- Press Copy above and paste the prompt.
- Run an "upscaler" (such as Ultimate SD Upscale) to enlarge it, then save.
- This one lives inside Photoshop, for people who already have it. Otherwise use Canva above — it's easier and also makes a print PDF.
- Make a New document: 24.25 × 36.25 inches, 300 DPI, CMYK Color.
- Select the whole canvas, press Copy above, and run Generative Fill with the prompt.
- Pick the version you like, then File → Save As → Photoshop PDF using the PDF/X-4 preset with bleed and marks turned on.
Tailored for each AI tool
Use the prompt above, then add these tool-specific settings. First pick a tool that can reach your final size — many image generators top out around 1–2K pixels (plenty for a card), though Nano Banana Pro/2 (up to 4K), Stable Diffusion, and Recraft (vector) can go large; for a 24 × 36 poster or bigger, lean on those, or upscale a smaller render. Two kinds of tools: a few export a real print PDF; most give a great RGB source image and we finish the prepress.
Adobe Firefly in Photoshop / Illustrator
The closest to a finished file — Firefly lives inside Adobe's print apps, so you control size, DPI and CMYK.
Canva Magic Media
Simple layouts with bleed built in. The AI image is RGB, but Canva's print export adds the bleed and crop marks.
ChatGPT GPT-4o / DALL·E
Great for backgrounds and imagery; weak at precise text. You can't set exact pixels, DPI or CMYK.
Nano Banana Google Gemini image
Gemini's image model — generates and edits, superb at refining an image and holding a subject consistent. The newer Nano Banana Pro (Gemini 3 Pro Image) and Nano Banana 2 (Gemini 3.1 Flash Image) output up to 4K (~4096 px) — big enough for a poster.
Midjourney
The highest-quality generated art of the image tools — still RGB, no crop marks.
Stable Diffusion ComfyUI · Leonardo
The one image path where you set exact pixels — so you can hit 300 DPI natively.
Ideogram
The strongest at rendering readable words inside the image — handy when your design has a headline or logo text baked in.
Recraft
Can generate vector art (SVG) that scales to any print size with no resolution limit — ideal for logos, icons and flat graphics.
Using Grok, Flux, Bing / Copilot or another generator? Treat it like the image tools above — match the aspect ratio, generate the largest size available, and send us the source. We'll finish the prepress.
What each setting means
Five things separate a screen graphic from a print file. Prompt for these and your AI art arrives clean.
Always the highest you can
Generate the most pixels your tool allows — up to 600 DPI. The viewing-distance minimum (~300 in hand, ~150 poster, ~100 banner) is the floor, never the target. More pixels is always better; it never hurts.
24 × 36 in → floor 3,638 px · max 14,550 px wideAdd 1/8″ all around
Printers cut through a stack of sheets, so color must run past the cut. Extend the background 0.125″ beyond every edge — a “bleed” — with no white border, or you risk a thin white sliver at the trim.
Trim 24 × 36 → build 24.25 × 36.25 inKeep text 1/4″ in
Cutting has a small tolerance. Hold text, logos and faces at least 0.25″ inside the trim so nothing important lands on the blade.
24 × 36 in → keep inside 23.5 × 35.5 inRGB looks brighter than it prints
Screens glow in RGB; presses lay down CMYK ink. Bright neon greens, blues and oranges are the ones that shift most. Pick rich, saturated-but-real color and expect a small settle on press.
Deep blacks print richest as a rich black (we build it C30 M60 Y30 K100). Small text stays 100% K so it prints sharp — you don't set this, just know it.
Lossless, not compressed
Export PNG or TIFF. Heavy JPEG compression bakes in blocky halos around edges and text that can't be removed later. If the tool offers a PDF, high-quality PDF is welcome too.
No hairlines, no tiny type
Lines thinner than a hair and text under ~6 pt can drop out or fill in on press. Keep strokes visible and type readable at arm's length.
Big natively, or upscale
Prefer a model that renders large natively — Nano Banana Pro / 2 at 4K (in Google AI Studio) or Stable Diffusion. For smaller tools, generate the largest they allow, then enlarge with an upscaler: a built-in “Upscale”, Topaz Gigapixel AI (paid), Upscayl (free) or Magnific AI. Or send us the biggest file and we'll upscale it.
Keep words out of the image
AI tools are notoriously bad at spelling text inside a picture. Keep important wording — names, prices, phone numbers — out of the AI image; add it in a layout tool, or send it to us as text and we'll set it cleanly.
Bleed, trim & safe zone
The honest part
Some things AI simply can't do — and we finish for free
No prompt makes a generative AI output true CMYK, printer crop marks, or spot colors. That's normal. Get the source art high-quality with the prompt above, and our prepress team does the rest.
You / your AI get the source right
- Correct size + 1/8″ bleed on the canvas
- 300 DPI / enough pixels, rendered large
- Text & logos inside the safe zone
- Bold color, solid blacks, no tiny detail
- Lossless PNG or TIFF export
PrintSimply finishes the prepress
- Convert RGB → CMYK and check the color
- Add crop marks & confirm the bleed
- Set rich black & keep small text 100% K
- Flatten, preflight & build the print PDF
- Send you a proof before anything runs