E-commerce-grade product images: reusable prompt structure and reference image usage

Centered on Nano Banana Pro's image-to-image capability, this demonstrates how to reliably produce e-commerce-grade product images using "reference image + structured prompt." The workflow first generates a real-world product reference (matte bottle, restrained lighting) from the production line, then submits it as image_input to the authoring line. With a prompt that explicitly describes material, light positions, background, composition, and aspect ratio, the model rearranges lighting and scene on the same reference image rather than re-guessing the product's appearance. The prompt approach has four layers: subject and material (bottle material, label text position), light positions and shadows (key light direction, softbox size, rim light), background and environment (seamless paper backdrop, contact shadow), and output constraints (2K resolution, 4:5 vertical composition, PNG). This structure suits batch layout trials for product hero images, brand social media assets, and advertising creatives; use the reference image to maintain product consistency, then iterate scenes with prompt variants.

Style Product Image · ecommerce · Studio lighting

Model Nano Banana Pro

Reference inputs

image_input

Try it yourself

Prompt

Try this model

Studio e-commerce product photography: a matte ceramic cosmetic bottle with a minimalist label, centered on a seamless warm gray background. A large softbox key light from the upper left creates soft gradient falloff, a subtle rim light on the right edge separates the bottle from the background, and a realistic contact shadow grounds the object firmly in the frame. Clean commercial product catalog composition, 4:5 vertical framing, sharp focus on the label, shallow depth of field, neutral color grading, no props, no text overlays, rich surface texture detail.

Copy the full prompt and adapt it for your own generation.