Prompt: Write a Product Photo Prompt

A structured prompt for AI product photography that specifies lighting, composition and background instead of leaving them to chance.

The prompt

Open in Studio

Generate a product photo with the following exact specification.

Subject: {{PRODUCT — be specific about colour, material and shape}}

Composition: {{ANGLE — e.g. three-quarter view, flat lay, close-up on texture}}, product filling {{APPROXIMATE FRAME PERCENTAGE — e.g. 60%}} of the frame.

Style: {{STYLE — e.g. clean studio product shot, lifestyle context, minimalist}}

Lighting: {{LIGHTING — e.g. soft diffused light from upper left, single hard shadow, golden hour}}

Constraints:
- Background: {{BACKGROUND — e.g. seamless white, blurred neutral surface}}
- No text, logos or watermarks anywhere in the image unless explicitly part of the product
- No visible people or hands unless specified above
- Aspect ratio: {{ASPECT RATIO — e.g. 1:1, 4:5, 16:9}}

Example: for online marketplace listings, a common brief is a 1:1 seamless white background shot with soft, even lighting and no shadow.

Return only the generated product image, with no accompanying text or caption.

Before generating, check that the composition, lighting and background above don't conflict with each other (e.g. a hard shadow requested alongside "soft diffused light").

Example

Input

PRODUCT — be specific about colour, material and shape: matte black ceramic pour-over coffee dripper with a spiral internal ridge
ANGLE — e.g. three-quarter view, flat lay, close-up on texture: three-quarter view
APPROXIMATE FRAME PERCENTAGE — e.g. 60%: 65%
STYLE — e.g. clean studio product shot, lifestyle context, minimalist: clean studio product shot
LIGHTING — e.g. soft diffused light from upper left, single hard shadow, golden hour: soft diffused light from upper left
BACKGROUND — e.g. seamless white, blurred neutral surface: seamless white
ASPECT RATIO — e.g. 1:1, 4:5, 16:9: 1:1

Output (a written description, not a generated image)

What a good result should look like: A matte black ceramic pour-over coffee dripper shown from a three-quarter angle, filling roughly two-thirds of a square 1:1 frame against a seamless white background. Soft, diffused light falls from the upper left, producing a gentle gradient across the dripper's curved body and a faint, soft-edged shadow to the lower right rather than a hard line. The ceramic surface shows a subtle matte sheen with visible texture variation, not a flat, glossy CGI-like finish. The spiral internal ridge is visible and in sharp focus, consistent with a clean studio shot intended for an online marketplace listing. No text, logos, watermarks, hands or other objects appear anywhere in the frame, and the white background is evenly lit with no visible seams or colour cast.

Why it works

Separating lighting direction from composition controls whether a shot reads as professional or flat, more precisely than a vague 'nice lighting' request.
Stating exclusions (no text, no hands) as constraints gets them followed more reliably than leaving them as an unstated assumption.
The pre-generation consistency check catches self-contradictory specs, a common cause of odd or uncanny image results.

Why the basic version fails

“A photo of my product on a white background” leaves lighting, angle and framing entirely up to chance — you get a technically correct but generically lit image indistinguishable from thousands of others, and often have to regenerate several times to land on something usable.

What this prompt fixes

Specifying lighting direction and quality separately from composition is what actually controls whether an image looks like a professional product shot or a flat snapshot. The consistency check catches self-contradictory specs (a common cause of odd, uncanny results) before generation instead of after.

Tips per model

Gemini responds well to precise lighting direction (“soft light from upper left”) rather than vague mood words (“nice lighting”).
ChatGPT’s image generation follows explicit exclusions (no text, no hands) more reliably when they’re stated as constraints rather than assumed.
If a first generation looks close but not right, change one field at a time (lighting, then angle, then background) rather than rewriting the whole prompt — it makes it much easier to tell which change actually fixed the result.