Packaging design prompts. Material is the brand.
Pick a product, brand vertical, form factor, material, print style, color, and setting; get a Midjourney, Nano Banana, or DALL-E-ready brief with negative prompt and finish callouts. Browser-only; nothing is saved.
Product, material, form, color, setting.
Your packaging prompt.
Click build to generate.
What good looks like.
Holloway: clean-beauty serum bottle on travertine plinth
Subject: 30 ml amber apothecary glass dropper bottle. Vertical: clean beauty. Print style: typography-led (brand name as the entire front, ingredient list on the back). Color: warm earthy duotone (cream label with terracotta brand name). Setting: studio plinth, travertine surface, single soft north-light source. Aspect: 4:5. The amber glass makes the product look medicinal and serious; the warm typography softens that into editorial. The travertine is visually neutral but adds material story. This is the canonical clean-beauty packaging photo and works in Midjourney v6.1 as well as DALL-E 3.
Northrun: specialty coffee bag, kraft paper, flat-lay with whole beans
Subject: 12 oz stand-up coffee pouch with one-way valve and tin-tie closure. Material: kraft paper (visible fiber, slight grain). Print style: typography-led monochrome black ink. Color: monochrome ink. Setting: flat-lay with a small spill of whole beans, a brass scoop, and a single coffee filter, all on a textured concrete surface. Aspect: 1:1 (Instagram square). The kraft paper signals craft and origin; the monochrome ink keeps the bag from competing with the bean detail; the flat-lay shows the buyer what they are getting. Works as the canonical specialty-coffee photo. Best paired with brand work via the brand identity service.
Quietwick: luxury candle in ceramic vessel with copper foil mark
Subject: 8 oz ceramic candle vessel (matte glaze, slight imperfection allowed) with a copper-foil-stamped brand mark on the front. Material: ceramic. Print style: foil-stamped accent (the brand mark catches light; the rest is matte). Color: single accent on neutral (cream ceramic with copper foil). Setting: studio plinth, dark moody backdrop, single side light. Aspect: 4:5. The ceramic gives weight and craft; the copper foil is the only flash; the dark background makes the foil pop. Foil-stamped print is the single hardest packaging detail to render in AI - test 4 to 6 variations and pick the one where the foil reads as foil rather than printed gold. Reference: Figma for the dieline and Anthropic tooling for the prompt iteration.
Packaging is a material story.
A packaging photo's first job is to tell the buyer what the product feels like. Material does that more than color, more than typography, more than the actual product photo. Kraft paper says craft; amber glass says serious; ceramic says heirloom; aluminum says modern; bioplastic says new. A prompt that names the material precisely - "amber apothecary glass with slight texture, light passes through with warmth" - gets a usable photo on the first generation. A prompt that says "premium glass bottle" gets a stock-photo-looking render with a plastic sheen. The named material is the brand.
Material is the brand
Buyers pick up a package before they read the label. The material communicates first: kraft paper says "craft and origin"; amber glass says "ingredient-led and clinical"; ceramic says "made by hand"; aluminum says "modern and sustainable"; recycled plastic says "honest about cost". The packaging prompt should name the material explicitly with one tactile detail (slight grain, slight imperfection, hand-thrown look). AI image models default to "premium plastic" if the material is not named, which is the worst possible default. Reference: web.dev material design articles for the underlying material psychology.
Form factor signals tier
Tall and slender reads as premium because it costs more in shelf space and ships less efficiently. Squat and wide reads as accessible. Cube reads as clean and modern. Cylinder reads as classic and ingredient-led. Asymmetric reads as boutique and design-forward. Buyers price-anchor on form factor before they look at the label - which is why luxury fragrances are tall and slender, supplements are squat (cheaper to ship), and coffee bags are pouches (cheaper to fill). The form factor is a brand decision, not just a manufacturing decision. The prompt should name the silhouette so the AI does not default to a stock cylinder.
Print style is not illustration
"Print style" describes how the brand mark and label copy meet the material. Typography-led packaging puts the words first; illustration-led puts a drawing first; photographic puts a hero image first; foil-stamped is a finish technique that lives on top of any of those. The most expensive-feeling packaging is typography-led plus a single foil-stamped accent because both signals are restraint. The cheapest-feeling is photographic plus aggressive color because both are over-explained. The prompt should pick one print style and stay there - mixed signals are what produce the "looks like every other AI render" effect. Pair this with the brand identity service for the underlying type and mark system.
Color reads on shelf
Most packaging gets evaluated next to competitor packaging on a retail shelf or in a feed of competitor product photos. The buyer's eye looks for difference and signal-of-tier in the same glance. Monochrome black on kraft paper stands out in a category where every competitor is using bright color (specialty coffee, for example). Bright editorial color stands out in a category where every competitor is using muted neutrals (clean beauty). The color choice should be made against the competitive set, not in isolation. Use the color contrast checker to verify the label copy passes 4.5:1 contrast against the chosen background.
Lifestyle vs studio choice
Studio plinth shots look professional, expensive, and slightly clinical. Lifestyle scenes look warm, real, and slightly amateur. The trade-off is brand-specific: clean beauty leans studio (the credibility cue is clinical), craft food leans lifestyle (the credibility cue is real-people-eat-this), specialty coffee splits the difference (studio for the bag itself, lifestyle for the brewing scene). The setting choice should match where the photo will be used: studio for the product page hero, lifestyle for the email and social, flat-lay for the launch announcement. Generate one of each from the same prompt by changing only the setting field.
Negative prompt as guard
The biggest AI packaging failure is text artifacts - warped typography, hallucinated extra labels, fake competitor logos in the background. The fix is a negative prompt block: "no text artifacts, no warped typography, no extra labels, no logo of a real competing brand". Without that, the model adds plausible-but-wrong text and you spend the next generation trying to fix it. Other useful negative prompts: "no human face" (keeps the product the subject), "no busy background" (keeps the product the subject), "material must look like real material at close range" (forces the model away from over-rendered plastic look). For Midjourney use --no, for Nano Banana use a "do not include" line, for DALL-E use plain language inside the prompt. Reference: Stability AI documentation for the underlying negative-prompt mechanics across diffusion models.
Related work: Brand identity for full packaging systems. App icon prompt for the digital counterpart. Website hero prompt for the matching landing page hero. Business name generator if the brand still needs naming. Web design for the product detail page that the packaging shot ships into.
Five answers.
Should I use Midjourney, Nano Banana, or DALL-E for packaging shots?
Each model handles different categories better. Midjourney v6.1 is strongest for studio-plinth shots with controlled lighting and is the default for clean beauty, fragrance, and luxury packaging. Nano Banana (Gemini's image model) is currently the strongest at lifestyle scenes because it preserves photographic realism in surrounding context. DALL-E 3 is the most flexible at handling unusual material specs (ceramic with hand-thrown imperfection, foil-stamped marks) but requires more iterations. None of the three reliably renders typography correctly on the label - always plan for a vector cleanup pass where you re-render the label in Figma before shipping the photo to a product page.
Why are AI-generated packaging photos useful before a real photoshoot?
A real photoshoot for a packaging line costs between $3,000 and $20,000 depending on the production team, location, and asset count. Generating 12 to 16 AI-rendered concept shots before the shoot lets the founder choose composition, lighting, and setting in advance - rather than discovering on set that the chosen plinth color does not work with the kraft paper. The AI shots are not the final assets; they are the brief for the photographer. This pre-visualization pass reduces both shoot time and revision rounds dramatically. For pre-launch direct-to-consumer brands without a photoshoot budget, the AI shots can sometimes serve as the launch assets, but plan for a real shoot within the first 90 days of the launch.
How do I keep typography from rendering wrong?
All three image models still struggle with typography on labels. Words come out warped, misspelled, or invented. The workaround is to generate the packaging shot without legible label text - use the negative prompt "no text artifacts, no warped typography" - and then composite the real label on top in Figma or Photoshop. Generate the shot with the label as a clean color block; render the actual brand name and copy in vector type; mask the type onto the bottle. This composite approach gives you AI-rendered material and lighting plus pixel-perfect typography. The all-AI single-pass approach is reserved for early concept generation where the typography accuracy does not matter yet.
What's the right aspect ratio for product packaging shots?
Depends on where the shot lives. 1:1 (Instagram square) is the most-shared aspect for DTC brands and is the default for the launch carousel slide 1. 4:5 (Instagram portrait) is the second-most-shared and gives more visual real estate; it is the right choice for the product page hero. 16:9 (web hero) is the right aspect for the homepage hero band but feels stretched on social. 3:4 (Pinterest) is the right aspect for Pinterest specifically, where vertical formats outrank square ones. Generate the same composition in 4:5 and 1:1 from the start so you have both for the launch.
Does this tool save my prompts?
No. Every value you enter and every prompt that gets built lives in your browser tab only. Nothing is transmitted to a server, stored in a database, or synced across devices. Close the tab and the data is gone. The Copy prompt button puts the assembled brief on your clipboard; that is the only output path.
Packaging is material.
Our brand-identity engagements pair a packaging system with the wordmark, label typography, dieline, and a per-SKU asset library. Two-week turnaround on the system, two-month timeline through real-photoshoot delivery.