TL;DR: A product photo prompt generator turns a product description into a structured image prompt: angle, surface, light, shadow, lens, framing. Below are 30 copy-paste templates organised by the shot you actually need. Use them for concept, mood and campaign work, and photograph the real product for anything a buyer will treat as the item itself.
Most pages that promise a product photo prompt generator hand you a paragraph of adjectives and an implied promise that you can stop shooting. This one is organised the way a shoot is organised, by shot type, and it starts with the part the others leave out: a generated image is a picture of something that resembles your product, not a picture of your product. That distinction decides which of these templates you can safely put in front of a buyer.
What is a product photo prompt generator, and what does it produce?
It produces text. A generator takes "matte black ceramic mug, 350ml, our bestseller" and returns a prompt that names the camera angle, the surface it sits on, the backdrop, the lighting setup, the shadow behaviour, the focal length, the crop and the things that must not appear. The image model does the rest.
That split matters commercially. Prompt Architects generates and stores the prompt; it does not render pictures. Say that plainly and the rest of this page makes sense: what you are buying from a prompt tool is repeatability across forty products, not pixels.
A good product prompt reads like a call sheet. A bad one reads like a mood board with 8k, ultra detailed, professional, award winning stapled to the end. Quality tags are the weakest instruction in the prompt because they describe how you want to feel about the result rather than anything the model can place in the frame. A prompt that names a real lighting setup and a real focal length gives the model something it can place in the frame; a stack of superlatives does not. That is our own observation from writing these, not a documented vendor behaviour.
Why is a generated product image not your product?
Because the model has never seen your product. It has seen a very large number of objects in that category, and when you ask for yours it composes a plausible member of the category. Plausible is the operative word. The handle will be a handle, but not your handle. The label will look like a label from three feet away and dissolve into approximate letterforms at full size. Stitch counts change. Button counts change. A four-port hub grows a fifth port. A logo becomes a logo-shaped smudge.
This is not a prompting failure you can write your way out of, and the vendors say so in their own documentation. OpenAI's limitations list for its GPT Image models states that "the model can still struggle with precise text placement and clarity" and that it "may have difficulty placing elements precisely in structured or layout-sensitive compositions" (developers.openai.com, image generation guide, accessed August 27, 2026). The same list warns that the model "may occasionally struggle to maintain visual consistency for recurring characters or brand elements across multiple generations." Text placement, layout precision and brand-element consistency are, between them, an exact description of the product-photography problem.
The practical line: a from-scratch generated image should not be a marketplace listing's main image. Not because a policy in this article says so, but because the image does not depict the item on offer, and everything downstream of that fact goes badly.
What should you use generated product images for, then?
Four jobs, all of them real, none of them "this is the item you will receive":
- Concept. Deciding what the shot should be before you book a studio day. Twelve generated angles cost minutes and settle an argument that would otherwise cost a reshoot.
- Mood and art direction. Establishing palette, light quality and styling so the photographer, the retoucher and the brand lead are describing the same picture.
- Backdrops and environments. Generating the scene the real product will later be composited into. The environment is invented, which is fine, because nobody is buying the table.
- Campaign and social exploration. Banner treatments, seasonal variants, story frames and hero layouts where the product is illustrative and the copy is doing the selling.
Everything on that list has one property in common: the generated pixels are context, not evidence.
Which workflow actually holds up for a real listing?
Start from a photograph of the real product and change everything except the product. That is image-to-image, and it is the only version of this that survives contact with a listing page, because the product in the frame was photographed rather than invented.
The shape is simple. Shoot the product once, properly, on any clean surface with any adequate light. Then use an editing or reference workflow to replace the background, restyle the surface, relight the scene or seasonalise the whole frame. The label stays your label because it is a photograph of your label. The port count stays right because nobody regenerated the ports.
Both major vendors document this path. Google's image-generation guide has a section called high-fidelity detail preservation whose instruction is that "To ensure critical details (like a face or logo) are preserved during an edit, describe them in great detail along with your edit request" (ai.google.dev, Gemini API image generation, accessed August 27, 2026). OpenAI documents a mask-based edit path and is candid about its limits: "Masking with GPT Image is entirely prompt-based. The model uses the mask as guidance, but may not follow its exact shape with complete precision" (developers.openai.com, accessed August 27, 2026). Guidance, not a hard boundary. Check the edges.
The full craft of that swap, including matching light direction between subject and new backdrop and getting contact shadows that do not read as pasted on, is its own subject: see background replacement prompting, which covers masked, reference-image and text-only variants tool by tool.
Three starting prompts for the workflow that holds up:
Using the attached photograph of [PRODUCT], replace only the background and the
surface. New setting: [SURFACE] under [LIGHT SETUP], background [BACKGROUND].
Keep the product itself completely unchanged: same shape, same proportions, same
colour, same label text, same finish, same wear. Match the new lighting direction
to the existing highlights on the product, and add a contact shadow consistent
with a key light from [DIRECTION]. Do not redraw, restyle or re-letter any part
of the product.
Using the attached photograph of [PRODUCT], keep the product pixel-accurate and
relight the scene only. Target look: [LIGHT SETUP], shadow [HARD/SOFT] falling
[DIRECTION], background [BACKGROUND]. Preserve the label exactly as photographed,
including [EXACT LABEL TEXT] rendered in the same typeface, size and position.
Preserve every [COUNTABLE FEATURE, e.g. button, port, stitch row]. Output
[ASPECT RATIO].
Using image 1 as the product and image 2 as the environment, place [PRODUCT] into
that scene at [POSITION IN FRAME], scaled so that it reads as [REAL WORLD SIZE].
The product from image 1 must remain completely unchanged in shape, colour and
markings. Adjust only shadow, reflection and colour cast so it sits in image 2's
light, which comes from [DIRECTION] at [COLOUR TEMPERATURE].
Which nine fields does every product photo prompt need?
The templates below are all built from the same nine fields. Learn the fields and you can write your own for any shot that is not listed.
| Field | What it controls | Example value |
|---|---|---|
| Product | Object, material, finish, colour, size | matte black stoneware mug, 350ml, unglazed foot |
| Angle | Camera height and rotation | elevated 45 degrees · straight-on eye level · top-down |
| Surface | What it sits on, and its texture | honed grey concrete · pale oak, visible grain |
| Background | What is behind, and how far back | continuous white sweep · blurred kitchen, 2m behind |
| Light | The real setup, not a mood word | large softbox camera-left at 45 degrees, white bounce right |
| Shadow | Direction, hardness, contact | soft shadow falling right, grounded contact point |
| Lens | Focal length and aperture | 100mm macro at f/11 · 50mm at f/2.8 |
| Framing | Crop, negative space, aspect ratio | centred, 30% margin, square · product left third, 16:9 |
| Exclusions | What must not appear | no text, no props, no reflections of the studio |
Two of those fields carry most of the difference between an amateur prompt and a professional one, and both have their own vocabulary. Lighting setups have real names with real definitions, and using them correctly is worth more than any number of quality tags: the 40-term lighting vocabulary is the reference. The same is true of glass: camera and lens terms explains why an 85mm at f/11 and a 24mm at f/2 produce different objects from the same description.
For the exclusions field, a short, specific list beats a long generic one. If you want a fuller treatment of what to exclude and which models honour a dedicated exclusion field at all, the negative prompt generator covers the support matrix.
The 30 templates, by shot
Replace every [BRACKETED] field. Delete any line that does not apply to your product rather than filling it with an adjective.
Pure white and e-commerce cutout
The catalogue workhorse. Product isolated, no context, nothing competing for attention. This is also the shot you should photograph rather than generate, for every reason in the section above.
A high-resolution studio product photograph of [PRODUCT], centred on a pure white
continuous background. Lighting: two large softboxes at 45 degrees either side plus
a white bounce card below, giving a near 1:1 key-to-fill ratio so shadows barely
register. Shadow: minimal, with a soft grounded contact shadow directly beneath
the product. Camera: 100mm at f/11, straight-on eye level. Framing: centred with
20% margin on all sides, square. No props, no text, no visible studio equipment.
A cutout-ready studio photograph of [PRODUCT] on a fully transparent background.
Lighting: broad frontal soft light with a subtle top rim to separate the upper edge.
Shadow: none. Camera: 85mm at f/11, [ANGLE]. Framing: product fills 80% of the
frame height, centred, square. Edges must be clean and unambiguous against the
background, with no spill, no halo and no cast shadow.
A studio packshot of [PRODUCT] on a pure white background, three-quarter view
rotated [15/30/45] degrees to the [LEFT/RIGHT] and elevated 20 degrees so the
[TOP FACE/OPENING] is visible. Lighting: single large softbox camera-left,
silver reflector camera-right, hair light above and behind for edge definition.
Shadow: soft, falling to the right, contact point sharp. Camera: 105mm at f/13.
Framing: square, 25% margin. No props, no reflections, no text.
Lifestyle in context
The product doing its job, in the place it does it. This is where generated backdrops earn their keep, because the room is invented and nobody is buying the room.
A natural lifestyle photograph of [PRODUCT] in use in [SETTING], mid-[ACTION].
Lighting: window light from camera-left, late morning, soft and directional, with
gentle fill from a white wall opposite. Shadow: soft, falling right, long enough to
read as low sun. Camera: 50mm at f/2.8, eye level, product in the near third and
sharp. Background: [SETTING DETAILS] falling into gentle blur. Framing: 4:5,
product occupying the left third. Nothing branded in the background.
An editorial lifestyle photograph of [PRODUCT] resting on [SURFACE] in [SETTING],
unattended, as if just put down. Lighting: overcast daylight through a large
window, flat and even. Shadow: soft and directionless. Camera: 35mm at f/4,
slightly above eye level looking down at 20 degrees. Palette: [2-3 COLOURS].
Framing: 3:2 landscape with generous empty space to the [LEFT/RIGHT]. No people,
no text, no competing products.
An aspirational lifestyle photograph of [PRODUCT] in [OUTDOOR LOCATION] during
golden hour. Lighting: low warm sun behind and to the right of the product,
creating a rim along its upper edge, with reflected fill from the ground. Shadow:
long, soft-edged, falling toward camera. Camera: 35mm at f/2.8, low angle 30cm
from the ground. Atmosphere: slight haze catching the backlight. Framing: 16:9,
horizon in the upper third. No people, no branding other than the product.
Hero and banner
Wide, deliberate, built around empty space where copy will go. Write the negative space into the prompt or the layout team will crop your product in half.
A wide hero banner photograph of [PRODUCT] positioned in the right third of the
frame against [BACKGROUND]. Lighting: single large softbox from camera-right at 60
degrees, subtle blue-grey fill from the left. Shadow: soft, falling left across the
surface. Camera: 85mm at f/8, eye level. Framing: 16:9, with the entire left half
clean and uncluttered for headline text. Colour palette: [2-3 COLOURS]. No text
rendered in the image.
A dramatic low-key hero photograph of [PRODUCT] on [SURFACE] against a near-black
background. Lighting: single hard key light from high camera-left at 60 degrees,
one narrow strip light behind to trace a bright edge along the product's right
side, no fill. Shadow: hard, deep, falling to the lower right. Camera: 100mm at
f/11. Framing: 21:9, product centred and small in frame, dark negative space
around it. No props, no text.
A clean evergreen hero photograph of [PRODUCT] floating slightly above [SURFACE]
against a smooth [COLOUR] gradient background. Lighting: broad even soft light from
above and front, near 1:1 key-to-fill. Shadow: a small soft ellipse directly below,
indicating height. Camera: 85mm at f/11, straight-on eye level. Framing: 16:9,
product centred, 35% margin. No seasonal props, no text, nothing that dates the
image.
Detail and texture macro
Close enough that the material is the subject. This is where a generated image most obviously stops being your product, so treat it as art direction unless you shot it.
An extreme macro photograph of the [SPECIFIC AREA] of [PRODUCT], filling the frame.
Lighting: raking side light from camera-left at a shallow 15-degree angle to
exaggerate surface texture, with a small silver reflector opposite. Shadow: fine,
directional, revealing every ridge. Camera: 100mm macro at f/8, focus stacked so the
full surface is sharp. Framing: square, abstract, edges of the product not visible.
Material: [MATERIAL AND FINISH].
A close-up construction-detail photograph showing [JOIN/SEAM/WELD/STITCH] on
[PRODUCT]. Lighting: soft diffused light from directly above plus a hard accent
from camera-right to define the edge. Shadow: tight and shallow. Camera: 100mm
macro at f/11, 30 degrees off perpendicular. Framing: 4:5, the detail occupying
the centre 60%. Show [SPECIFIC COUNTABLE DETAIL] clearly. No text, no props.
A material-contrast macro photograph where [MATERIAL A] meets [MATERIAL B] on
[PRODUCT]. Lighting: large soft source above and behind at 120 degrees, so one
material catches specular highlight and the other stays matte. Shadow: soft, thin,
along the join. Camera: 100mm macro at f/9. Framing: square, diagonal composition
running lower-left to upper-right. Emphasise the difference in reflectivity between
the two surfaces.
Scale reference
The single most under-shot slot in e-commerce, and the one that quietly causes returns. Say what the reference object is and what size it implies.
A product photograph of [PRODUCT] held in an adult hand against a plain [COLOUR]
background, to show scale. The hand is [SKIN TONE], relaxed, entering from the
lower right, with the product held [GRIP DESCRIPTION]. Lighting: large softbox
camera-left, white bounce right, near 1:1 ratio. Shadow: soft, minimal. Camera:
85mm at f/5.6, eye level. Framing: 4:5, hand and product filling the centre.
The product measures approximately [DIMENSIONS]. No jewellery, no nail polish,
no text.
A scale-reference product photograph of [PRODUCT] beside [COMMON OBJECT, e.g. a
standard credit card / a 330ml can] on [SURFACE]. Both objects sharp and in the
same focal plane. Lighting: even soft light from above and front. Shadow: soft,
falling directly back. Camera: 100mm at f/11, straight-on eye level so relative
sizes are undistorted. Framing: 3:2 landscape, both objects centred with equal
margin. No perspective exaggeration.
An environmental scale photograph of [PRODUCT] installed or placed in [REAL
SETTING], showing its size relative to [FIXED FEATURE, e.g. a doorway / a
kitchen counter]. Lighting: available daylight, soft, from camera-left. Shadow:
natural and consistent with a single window. Camera: 35mm at f/5.6, eye level,
straight-on to minimise distortion. Framing: 3:2, product in the centre third.
Flat lay
Top-down, graphic, and unforgiving about alignment. Say "perfectly perpendicular" or you will get a 15-degree tilt you did not ask for.
A top-down flat lay photograph of [PRODUCT] on [SURFACE], camera perfectly
perpendicular to the surface with no tilt. Lighting: large diffused source directly
overhead, plus low fill from all sides. Shadow: very soft, falling evenly outward.
Camera: 50mm at f/8. Framing: square, product centred, 25% margin. Palette:
[2-3 COLOURS]. No props, no text, no hands.
A styled top-down flat lay of [PRODUCT] surrounded by [3-4 PROPS] arranged on
[SURFACE]. Camera perfectly perpendicular. Composition: product in the centre,
props radiating outward with even negative space between every element and no
overlaps. Lighting: soft overhead daylight from a large window. Shadow: soft,
short, all falling in the same direction. Camera: 50mm at f/8. Framing: square.
Palette: [2-3 COLOURS]. No text.
A what-is-in-the-box flat lay showing [PRODUCT] with [ACCESSORY 1], [ACCESSORY 2]
and [ACCESSORY 3] laid out in a grid on a plain [COLOUR] surface. Camera perfectly
perpendicular. Every item fully visible, none overlapping, equal spacing. Lighting:
flat even overhead light, near 1:1 key-to-fill. Shadow: minimal. Camera: 50mm at
f/11. Framing: 4:5, generous outer margin. No text, no labels, no packaging copy.
On-model and in-hand
The moment the product stops being an object and starts being a use case. Say what the model is doing, not just that a model exists.
A full-body e-commerce photograph of a [MODEL DESCRIPTION] wearing [PRODUCT] in
[COLOUR], standing in a relaxed neutral pose against a plain [COLOUR] background.
Lighting: large softbox camera-left at 45 degrees, white bounce right, near 1:1
ratio. Shadow: soft, falling right. Camera: 85mm at f/5.6, eye level, full body in
frame with headroom. Framing: 4:5. The garment fits correctly and hangs naturally.
No visible branding other than the product. No text.
An in-hand demonstration photograph of [PRODUCT] being used to [SPECIFIC ACTION],
hands only, entering the frame from the [DIRECTION]. Lighting: soft directional
light from camera-left, gentle fill right. Shadow: soft, short. Camera: 50mm at
f/4, over-the-shoulder point of view looking down at 45 degrees. Framing: 4:5,
hands and product filling the centre, [SETTING] blurred behind. The action reads
clearly at thumbnail size.
A cropped worn-detail photograph of [PRODUCT] on the [BODY PART] of a
[MODEL DESCRIPTION], showing how it sits and moves. Lighting: soft window light
from camera-right, natural fill. Shadow: soft, following the contour. Camera:
85mm at f/2.8, focus on [SPECIFIC DETAIL], background falling away. Framing:
square, tight crop, face not in frame. Show the [FASTENING/TEXTURE/DRAPE] clearly.
Packaging
The box is a product too, and it is the shot most likely to expose a model's weakness with small text. Photograph the artwork; generate the environment.
A studio photograph of [PRODUCT] beside its [BOX/POUCH/TIN] on [SURFACE], both
sharp, package angled 30 degrees to show the front and one side face. Lighting:
large softbox camera-left, silver fill right, subtle top light on the package.
Shadow: soft, falling right, both objects grounded. Camera: 85mm at f/11, eye
level slightly elevated. Framing: 3:2, package left, product right, even spacing.
Leave the package artwork blank and unlettered.
An unboxing-moment photograph of a partially opened [BOX] revealing [PRODUCT]
inside, lid lifted and resting at an angle. Lighting: soft overhead daylight with
a warm bounce from the front so the interior is visible, not black. Shadow: soft,
short, falling back. Camera: 50mm at f/4, looking down at 60 degrees. Framing:
4:5, box filling the centre, [SURFACE] visible around it. Tissue and inserts
visible. No printed text on any surface.
A retail-shelf photograph of [PRODUCT] in packaging standing among [CATEGORY]
products on a [SHELF DESCRIPTION], the subject product in sharp focus and the
neighbours blurred. Lighting: overhead retail lighting, slightly cool, even.
Shadow: short, directly beneath. Camera: 50mm at f/2.8, eye level, straight-on.
Framing: 3:2, subject product in the centre third at correct shelf scale.
Neighbouring packaging unbranded and unreadable.
Group and range
Where consistency stops being a nice-to-have. Fix camera, light and spacing in the prompt and vary only the thing that is meant to vary.
A studio photograph of [N] colourways of [PRODUCT] arranged in a single straight
row on [SURFACE] against a plain [COLOUR] background, evenly spaced, all at
identical scale and identical rotation. Colours left to right: [COLOUR LIST].
Lighting: broad even soft light from above and front, near 1:1 ratio, identical
across every unit. Shadow: soft, short, identical direction under each. Camera:
100mm at f/13, straight-on eye level. Framing: 16:9, equal margins.
A size-range photograph of [PRODUCT] in [N] sizes, arranged smallest to largest
from left to right on [SURFACE], all in the same finish and rotation, with true
relative proportions preserved. Lighting: large softbox above and slightly front.
Shadow: soft, all falling in the same direction. Camera: 100mm at f/13,
straight-on eye level to avoid perspective distortion between units. Framing: 3:2,
even spacing. No text, no props.
A collection photograph of the full [RANGE NAME] range grouped on [SURFACE]:
[ITEM 1], [ITEM 2], [ITEM 3] and [ITEM 4], arranged with varied heights and slight
overlap so the group reads as one composition. Lighting: single large softbox
camera-left at 45 degrees, white bounce right. Shadow: soft, consistent, all
falling right. Camera: 85mm at f/11, elevated 20 degrees. Framing: 4:5. Palette:
[2-3 COLOURS]. No text.
Seasonal and campaign variants
The safest place for a purely generated image, because the point is the mood, not the merchandise. Keep the product treatment fixed and vary only the campaign layer.
A seasonal campaign photograph of [PRODUCT] in a [SEASON] setting: [SEASONAL
ELEMENTS] surrounding but not obscuring the product. Lighting: [SEASONAL LIGHT,
e.g. low warm winter sun from camera-right / flat cool overcast]. Shadow:
[LENGTH AND DIRECTION]. Camera: 50mm at f/4, eye level. Framing: 4:5, product in
the lower third with space above for a headline. Palette: [2-3 SEASONAL COLOURS].
No text, no dated references, nothing religious or region-specific.
A campaign restyle of the same [PRODUCT] hero shot, keeping camera, angle,
lighting direction and framing identical to the base image, and changing only the
palette and surface. New palette: [2-3 COLOURS]. New surface: [SURFACE]. Camera:
85mm at f/8, straight-on eye level. Framing: 16:9, product right third, clean left
half. Everything about the product itself stays identical to the base shot.
A gifting-context campaign photograph of [PRODUCT] presented on [SURFACE] with
[RIBBON/WRAPPING/GIFT ELEMENT], styled for a generic celebration rather than any
named holiday. Lighting: warm soft light from camera-left at 45 degrees, gentle
fill. Shadow: soft, short, falling right. Camera: 50mm at f/2.8, looking down at
30 degrees. Framing: square, product centred. Palette: [2-3 COLOURS]. No text,
no dates, no holiday-specific iconography.
Which shot belongs in which slot?
Not every shot is appropriate everywhere, and the difference is mostly about whether the viewer is treating the image as evidence or as atmosphere.
| Feature | Listing main image | Gallery / PDP | Ads and social | Email and banners |
|---|---|---|---|---|
| White cutout | Photograph it | Photograph it | Optional | Rarely used |
| Detail macro | Not this slot | Photograph it | Either | Either |
| Scale reference | Not this slot | Photograph it | Useful | Rarely used |
| Packaging | Not this slot | Photograph it | Either | Either |
| Lifestyle | Not this slot | Composite | Composite or generated | Generated is fine |
| Hero / banner | Not this slot | Optional | Generated is fine | Generated is fine |
| Flat lay | Not this slot | Composite | Either | Generated is fine |
| Seasonal variant | Not this slot | Rarely used | Generated is fine | Generated is fine |
The pattern is consistent: the closer an image sits to the buy button, the more of it needs to be a photograph.
What goes wrong most often, and how do you fix it?
The label is unreadable or invented. Expected, per the vendor limitations quoted above. Fix: photograph the label and edit around it, or leave the artwork blank in the prompt and add real artwork in layout. Never let a model write your ingredients list. If the shot genuinely needs rendered text, the GPT Image 2 prompt generator covers how to specify an exact string, but specifying it precisely is not the same as getting it back intact.
The product floats. No contact shadow, so it reads as a cutout dropped on a backdrop. Fix: always fill the shadow field, and name the contact point explicitly.
Perspective looks wrong on a small object. Almost always a focal-length problem. A wide lens close to a near object stretches whatever is nearest the camera. Fix: name 85mm to 105mm for packshots and keep wide focal lengths for environmental frames.
The range shot is inconsistent. Each unit lit and angled slightly differently, which no retouch will rescue. Fix: state "identical scale and identical rotation" and "identical lighting across every unit" in the prompt, and generate the row as one image rather than compositing separate generations.
Counts drift. Ports, buttons, stitches, pockets. This is the visual-consistency limitation quoted earlier showing up as arithmetic. Fix: photograph anything countable, and treat a generated count as decoration.
Where this fits
Everything above is text you paste somewhere else. Prompt Architects generates, structures and stores that text: it does not render images, and any page that implies a prompt tool will hand you finished photography is selling you the wrong thing.
What it does solve is the part that breaks at volume. The nine-field skeleton lives as a saved template, with {{PRODUCT}}, {{SURFACE}} and {{PALETTE}} as variables you fill per job instead of retyping the sentence and quietly dropping the exclusions line on product thirty-one. On plans, plainly: image prompt generation sits on Pro, Advanced and Team, verified on our own pricing page on August 27, 2026. The Free plan exists and includes "5 prompt enhancements per day, forever" according to our FAQ page, checked the same day, which is enough to run one template through it and judge the output before spending anything.
And the one line worth carrying out of this page: generate the scene, photograph the product.
Stop rewriting prompts. Start shipping.
Works with ChatGPT, Claude, Gemini, Grok, Midjourney, Ideogram, Veo3 & Kling. 5.0★ on the Chrome Web Store.
Create An Account