AI Prompts for Product Photography That Looks Like a Studio Shoot

Most people type "product photo of [thing]" into ChatGPT or Nano Banana and get back something that looks like a 3D render floating in a white void. Two problems are happening at once. First, the AI has never seen your product, so it invents one that looks nothing like what you sell. The fix there is reference images: upload a photo of your actual product, even a phone shot, and let the AI build around it. Second, people leave out the details that any real photographer would ask about before picking up a camera. Lighting direction, camera angle, background surface, composition, mood. These feel obvious in hindsight, but most AI image prompts skip all of them and let the model guess. It guesses badly.

The prompts in this post solve both problems: they start with your real product as a reference and give the AI explicit instructions for everything a studio photographer would control. They're written for the 2026 image stack — Nano Banana, GPT Image, and Flux — and pulled directly from the same library we use inside MarketPrompter's Ad Studio, where the Reference-to-Ad flow turns a reference photo plus a scene description into a finished ad.

The Five Elements Every Product Photo Prompt Needs

Every prompt below is built on the same five elements. If a prompt is producing weak output, almost always one of these is missing.

1. Cinematography — lens, focal length, depth of field, camera angle. "85mm, shallow depth of field, slight low angle" is doing more work than "good photo."
2. Subject — your actual product, locked in via a reference image. Never describe the product in words alone if you have a photo of it.
3. Action — what the product is doing or how it sits in the scene. Resting on a surface, partially submerged, mid-pour, tilted at an angle, in someone's hand.
4. Context — the environment around the product. Surface material, props, time of day, location. This is where most prompts get lazy.
5. Style & ambiance — color palette, lighting quality, mood, reference aesthetic ("editorial fragrance ad," "1950s diner," "Kodak Ektar color science").

You'll see all five in every prompt below. When you adapt them, swap the specifics — never delete the categories.

Which AI Image Model for Which Job (2026)

The model matters less than the prompt, but it does matter. Quick map:

• Nano Banana / Nano Banana Pro (Gemini) — best at respecting a reference image and at rendering legible text inside the image. Default for any product shot that needs your real product + a styled scene, especially when text or badges sit in the composition.
• GPT Image — strongest typography of any general-purpose model. Use when the image is essentially a poster: big words, brand colors, tight layout. Also excellent for in-image text overlays on product shots.
• Flux (Pro / Ultra) — most photoreal lighting and material rendering. Use for hero shots where the product needs to look like it came from a real camera, or for moody editorial sets.
• One-shot edits — when you need to take this product photo and put it in this scene, change the background, swap a color, or add a prop, both GPT Image and Nano Banana accept the reference and apply the change in a single generation. Lower friction than building a new generation from scratch.

Start With a Reference Image (Or a Sketch)

Without a reference image, AI image generators invent your product from scratch and get it wrong. Color, shape, label placement, proportions: all of it will be off. Attaching even a basic phone photo solves this because the AI uses your actual product as the subject and only changes the scene around it. If the product doesn't physically exist yet, a reference sketch works almost as well — use our free sketch to PNG pad to draw a rough shape and export a clean PNG the model can read. Tighten the crop on the product before you upload; loose backgrounds in the reference confuse the model.

How to Prompt AI for Studio-Style Product Photos

[ATTACH: a photo of your actual product — phone photo is fine, clear angle, decent lighting]

Using the product in the attached reference image, generate a studio-style product photograph.

Maintain the exact product from the reference. Only change the environment and lighting:
- Background: [white / colored gradient / textured surface — specify]
- Lighting: [soft diffused from upper left / hard directional from right / backlit with rim light / even studio lighting]
- Camera angle: [straight on / 45-degree / overhead flat lay / low angle looking up]
- Composition: [centered / rule of thirds / negative space on left for text overlay]
- Specific detail: [e.g., "show the texture of the fabric," "capture the transparency of the bottle," "lid slightly open to show contents"]

The product should sit on or interact with the surface, not float. Shadows should be visible and directional to ground it in the scene.

Why It Works

This prompt forces a decision on each of the five elements. The reference image handles subject. The bracketed fields cover cinematography (camera angle), context (background, surface), action (sit / interact / lid open), and style & ambiance (lighting quality and direction). The closing line about shadows is there because the most common AI-product-photo failure is a product floating with no contact shadow, which immediately reads as fake.

What This Looks Like in Practice

You take a phone photo of your candle on your kitchen counter. Nothing fancy. You attach it and run the prompt with: grey concrete surface, soft diffused light from upper left, 45-degree angle, rule of thirds. The AI keeps your exact product but replaces your kitchen with a studio setup. Same candle, same cork lid, same amber glass, but now it's lit properly, composed intentionally, and grounded with a real shadow. The reference image did the heavy lifting on accuracy. The prompt details did the heavy lifting on quality.

[!cta] heading=Try it with your product | text=1 free Ad Studio preview per month. No card required. | button=Open Ad Studio | href=/tools/ad-studio

Seven Copy-Paste Prompts for Real Product Sets

Every prompt below is built around [product] from a reference image. Paste the prompt, attach your product photo, swap [product] for your product noun (e.g., "skincare serum bottle," "running shoe," "cold brew can") and run.

Lifestyle Sets

Workshop bench, warm task light — artisan, craft, tools, men's grooming, premium home goods.

Place the [product] from the reference image into a wide shot commercial-style workshop set. Well-used woodworking or maker's workshop with warm overhead task lighting and natural side light from a dusty window. The [product] sits on a thick wooden workbench as the hero surface, slightly off-center. Subtle maker props nearby: wood shavings curled on the bench, a hand plane, a leather tool roll partially unrolled, a mason jar holding pencils. Background shows shelving with jars and tools, softly out of focus. Artisan brand shoot feel: warm, textured, authentic craft energy, shallow depth of field keeping the [product] tack-sharp against the rich workshop environment.

Forest morning, log hero surface — outdoor / hunting / adventure brands.

Place the [product] from the reference image into a wide shot commercial-style outdoor set. The setting is a dense forest with soft, natural morning light filtering through the trees. In the foreground, place a cleanly chopped wooden log positioned as the main focus and hero surface for the [product]. Surround the area subtly with hunting-themed props — camo fabric, binoculars, a thermos, a backpack — arranged neatly but not distracting from the log and [product]. Polished commercial setup: crisp details, balanced lighting, rich forest colors, gentle depth of field.

Gym floor, fitness brand campaign — sports / supplements / apparel.

Place the [product] from the reference image into a wide shot commercial-style gym set. Clean, modern gym with rubber flooring and minimal equipment visible in the soft background. The [product] sits on a flat rubber surface or low wooden bench as the hero element. Subtle fitness props nearby: a small gym towel draped casually, a water bottle, a resistance band coiled neatly. Overhead industrial lighting with warm tones, slight sheen on the floor. Fitness brand campaign feel: energetic but controlled, crisp product detail, shallow depth of field keeping the background soft.

Studio / E-commerce Hero

Pure white seamless — clean hero shot for product pages and marketplaces.

Place the [product] from the reference image into a wide shot commercial-style studio set. Pure, minimal studio environment: seamless white backdrop with the faintest warm cream gradient, a single smooth white surface for the [product] to sit on. No props. One soft shadow beneath the product, cast by a single large softbox from upper-left. Premium e-commerce hero shot: perfectly even lighting, the [product] as the only element, maximum negative space, tack-sharp focus.

Warm earth tones — organic skincare, wellness, candles.

Place the [product] from the reference image into a wide shot commercial-style set built entirely from warm, earthy materials: a terracotta tile surface, a linen cloth draped and creased naturally, dried grasses or pampas in a simple clay vase just at the edge of frame. Warm side lighting from camera right, golden and soft, casting long gentle shadows. Palette: terracotta, sand, cream, and muted olive. The [product] sits on the linen as the hero element. Organic skincare / wellness brand feel: warm, grounded, textured. Shallow depth of field with background elements soft.

Creative / High-Concept

Floating in color — tech launches and modern DTC hero images.

Place the [product] from the reference image into a creative commercial set where the product floats in mid-air against a smooth, saturated color gradient background (choose a gradient that complements the product's colors). No surface, no shadow beneath. The [product] is slightly tilted at a dynamic angle, as if frozen mid-rotation. Small abstract shapes — soft spheres, metallic rings, geometric fragments — float around the product at various distances, all softly out of focus. Clean studio lighting, product perfectly lit from two sides. Tech product launch / modern DTC hero image feel: bold color, confident floating composition, editorial precision.

Paint splash, frozen mid-air — energetic brand launch, beauty, color cosmetics.

Place the [product] from the reference image into a commercial-style set with art-studio energy. The [product] sits on a clean white surface, untouched and pristine. Around and behind it, vivid paint splatters in 3-4 bold colors (choose colors that complement the product's palette) arc across the white backdrop and surface — frozen mid-splash, dynamic and sculptural. The paint never touches the product. It frames it. Sharp studio lighting from above, high contrast between the clean product and the chaotic color around it. The product is calm at the center of creative chaos.

Surprise me — when the obvious scene is boring. Use this when you want a scroll-stopper that has no expected tie to the product category.

Place the [product] from the reference image into a surprising, visually arresting scene that has no obvious connection to the product's category but creates a compelling image anyway. Think editorial fashion spreads that put perfume in a diner, or sneakers in a baroque painting. Choose an unexpected but visually striking camera setup — an angle, lens, or lighting approach that would make a creative director pause mid-scroll — and commit fully to it. The scene should be specific and vivid, not random for the sake of random. It should photograph beautifully and make the [product] impossible to ignore. Unexpected, confident, "why does this work so well" energy.

If you want a one-click version of this workflow — paste a reference URL or upload a product photo, pick a scene, and get back a finished ad — that's exactly what Ad Studio's Reference-to-Ad does. The full prompt library these examples come from lives in the Prompt Library.

Generating Seasonal Product Photography with AI

[ATTACH: a photo of your actual product]

Using the product in the attached reference image, generate a grid with 4 seasonal variations. The product stays exactly the same. Only the environment, lighting, and color temperature change.

Spring: fresh greenery, soft natural light, pastel tones in background, linen surface
Summer: bright direct sunlight, outdoor setting, vibrant warm colors, condensation if product is cold
Fall: warm amber tones, textured wool or linen backdrop, moody directional light from one side
Winter: cool blue-white light, minimalist clean setting, matte surfaces, soft shadows

Rules:
- The product from the reference should look identical across all four
- Each season should feel distinct through color temperature and mood, not props
- Describe what you want in the scene, not what you don't want
- Avoid cliches (no pumpkins for fall, no snowflakes for winter) unless the product actually relates to them

Four seasonal photoshoots isn't realistic on a small-business budget. One reference image plus this prompt gets you a year of campaign visuals that all feature the same physical product. The "describe what you want, not what you don't want" rule matters because AI image generators respond to positive framing. Saying "no pumpkins" might still produce pumpkins.

AI Product Photos Sized for Instagram, Stories, and Ads

No model generates all three platform formats from a single prompt. Run each one separately, keep the lighting, style, and mood consistent, change only composition and aspect ratio.

Instagram feed (4:5): centered composition, breathing room around product, same lighting and background as your other versions, product fully visible.

Story / Reel cover (9:16): product in the lower third, clean negative space in the upper portion, same lighting and background, product fully visible.

Meta feed ad (4:5): product on one side, open space on the other for the headline, high contrast so it reads at small sizes in the mobile feed.

Cropping one image into three formats always looks like a crop. Running each format as its own generation with the same reference image and the same style language produces three images that look like they belong together but are each composed for how people actually view them on that platform.

Adding Text to Product Images (Don't Send It to Figma First)

The old advice was: generate the image clean, then add text in Figma or Photoshop. That advice is dead. GPT Image and Nano Banana Pro now render typography cleaner than most freelance designers, and they can build the design system around the text — frames, badges, rules, ornaments — in the same pass.

The rule: don't just describe the text. Describe the design system around it. Badges, underlines, frames, repeated motifs, supporting marks. That's what makes prompted text look designed instead of stamped on.

Examples of the right level of detail:

• "...across the top of the frame, the headline rendered in a chunky retro display style, slightly distressed, with a thin underline that curls into a small icon on the right end."
• "...playful hand-drawn type for the headline arcs over the product, with three tiny star bursts and a dotted halo framing the words. Two-color palette only."
• "...bold condensed type across the bottom band for the headline, paired with a small badge-style seal to the left and two thin horizontal rules above and below — text should feel like it belongs to a label system, not floated on top of the image."
• "...the headline set in a soft serif at the upper-left, with a hand-drawn arrow connecting it to the product and a small repeated motif (tiny stars, dots, or asterisks) anchoring the lower corners."

Notice what these examples do not do: they don't specify a font name, an exact color, or pixel coordinates. They describe the feel and the supporting design language and let the model pick the typography that fits. Over-specifying typeface and weight tends to produce stiff, generic results; under-describing the design system around the text produces text that looks pasted on.

Copy-Paste Text Block (Append to Any Prompt Above)

This block lives at the end of any product photo prompt when you want the image to include headline text or badges. It works filled out or left as-is.

Add ad copy and design elements that fit the product and scene. If no details are included in any of the fields below, take full editorial freedom. Headline: [text]. Supporting detail (optional): [text]. CTA: [text].

When you still drop into Figma: exact brand-font matches the model can't reproduce (e.g., a licensed Helvetica Neue 75 Bold), legal or regulatory text that has to be pixel-accurate, or when a client supplies a locked brand kit. Workflow in that case: generate the image clean, drop the background using our background remover, pull it into Figma, set the type there.

For everything else, prompt the text directly. It's faster, it composes with the lighting, and it doesn't look like two different applications stitched together.

Common Mistakes (Straight From the Course)

The same three mistakes show up in 90% of weak AI product photos:

1. No reference image. You typed "a sleek matte black water bottle on a desk" and got a generic stock-looking bottle. Always attach your real product. Even a phone photo.
2. Five elements collapsed into one vague line. "Nice lighting, cool background" isn't a prompt. Spell out cinematography, subject, action, context, and style/ambiance separately.
3. Negation prompting. "No clutter, no pumpkins, no people" tends to add the thing you're trying to remove. Describe the positive scene instead: "clean surface with two props," "autumn backdrop using rust and amber tones," "hands and props only, no full human figure."

When AI Product Photography Works (and When It Doesn't)

Use these when: you have at least one decent photo of your product (even a phone photo works) and need professional-looking images for your site, social, or ads without booking a studio.

Don't use these when: you have no photo of your product yet. Text-only descriptions won't produce an accurate version of your specific product. Take a few phone photos first. Also skip AI for products where fine details matter at zoom level (jewelry, electronics with ports, food plating). A real photo is worth the cost for those.

Keep in mind: AI-generated product photos work best for lifestyle contexts, seasonal campaigns, ad creative, and social formats where you need volume. Your hero product image on your listing or product page should still be a real photo if possible. Use AI to build the supporting content around it.

[!cta] heading=Run your first ad free | text=Upload a product photo or use a reference image. | button=Start free preview | href=/tools/ad-studio

---

Every prompt in this post — and a few hundred more — lives in the MarketPrompter Prompt Library. The deeper "how to prompt visuals" curriculum is in Module 3: Creative Strategy. And if you want the one-click version, paste a reference into Ad Studio.

---

Related Articles

• AI Prompts for Product Descriptions That Actually Convert
• AI Prompts for Ad Copy Across Every Platform
• AI Prompts for Landing Page Copy That Converts
• AI Character Prompts: Mascots, Sheets, and Characters You Can Actually Reuse