Nano Banana 2 for Product Photography: Prompts That Actually Sell

Product photography is expensive, slow, and requires a human with a camera, a studio, and opinions about lighting ratios. Nano Banana 2 — Promptyze’s name for Google’s Gemini Flash Image generation layer — cuts most of that out. You describe a product, you get a shot. Describe it well, you get a shot that looks like it belongs on a premium brand’s homepage rather than a stock photo graveyard.

This tutorial is about doing exactly that: writing prompts that produce clean, usable, commercially credible product images. No prior photography knowledge required. A basic understanding of what your product looks like: mandatory.

What You’ll Achieve

By the end of this guide, you’ll have a repeatable workflow for generating e-commerce product images across multiple formats — hero shots, lifestyle context, flat lay, detail close-ups, and social media crops. You’ll also understand why certain prompt structures work better than others, and how to dial in consistency across a product line so your shop doesn’t look like six different brands sharing a website.

What You Need

Access to Nano Banana through any of its available channels: the Gemini app at gemini.google.com, Google AI Studio at aistudio.google.com, the Gemini API via direct calls, or Vertex AI if you’re working at enterprise scale. The Gemini app is the fastest way to experiment. AI Studio gives you more control over generation parameters. The API and Vertex AI are for when you’re building a pipeline that generates hundreds of product variants without clicking a single button.

You also need a clear mental image of your product — color, material, rough dimensions, what problem it solves. The AI generator doesn’t know your specific product, so you are its eyes. Feed it vague information and it will invent something vague back at you.

Note 💡

Images generated through Nano Banana carry SynthID watermarks — Google’s invisible digital signature embedded in the pixels. These watermarks survive most editing and compression. For commercial use, factor this into your workflow and verify your platform’s policies on AI-generated product imagery.

The Anatomy of a Good Product Photography Prompt

A product shot prompt has four layers: the subject, the environment, the technical specs, and the mood. Miss any of them and you’ll get something that looks like it was generated — which is exactly what you’re trying to avoid. Here’s the structure that consistently produces clean results:

Subject → what the product is, its material, color, and finish. Environment → surface it sits on, background, props if any. Technical → shot type, angle, lighting style, resolution cue. Mood → the brand feeling — minimal, warm, clinical, editorial.

Let’s build from there.

Core E-Commerce Prompts: Copy, Paste, Generate

These prompts are written for specific product categories and shot types. Each one explains what it does and why the wording earns its place.

1. Clean white background hero shot — the workhorse of any product catalog. This is the Amazon listing, the Shopify thumbnail, the format every marketplace actually requires.

Matte black ceramic coffee mug, product photography, isolated on pure white background, soft diffused studio lighting, subtle shadow beneath the mug, shot from a 3/4 angle, crisp detail on texture, photorealistic, commercial product photo, 4K

The “soft diffused studio lighting” instruction prevents harsh shadows that flatten form. The “subtle shadow beneath” keeps the object from floating — a classic giveaway of cheap compositing. The “3/4 angle” is the standard e-commerce angle because it shows front, side, and top in one frame. Swap “matte black ceramic” for your actual material and color.

2. Lifestyle context shot — puts the product in a scene that tells a story. Converts better than white background for social ads and brand pages.

Matte black ceramic coffee mug on a weathered oak kitchen counter, morning light coming through a window, steam rising from the mug, fresh newspaper beside it, warm golden hour atmosphere, lifestyle product photography, shallow depth of field, photorealistic

“Steam rising” signals use, not just ownership — an important psychological cue in lifestyle shots. “Shallow depth of field” softens the background so the product stays dominant. The specific surface (“weathered oak”) and props (“fresh newspaper”) give the generator enough to build a coherent scene rather than inventing random clutter.

3. Flat lay for social media — overhead, graphic, works perfectly as an Instagram square or Pinterest pin.

Skincare serum bottle, flat lay product photography, top-down overhead shot, arranged on white linen fabric, surrounded by fresh rosemary sprigs and small white stones, even soft natural lighting, no shadows, clean minimal composition, photorealistic, square format

“Top-down overhead shot” explicitly sets the camera position — without this, the generator often defaults to eye-level. “No shadows” keeps the flat lay graphic and clean. The prop selection (rosemary, white stones) signals the brand category without you having to say “this is a natural skincare brand.” Adjust props to match your actual product’s positioning.

4. Detail close-up shot — shows material quality, texture, craftsmanship. Essential for jewelry, textiles, premium goods.

Close-up macro product photography of a leather wallet, dark tan full-grain leather, showing stitching detail and grain texture, single-source directional light from the left, deep shadows revealing texture, extreme shallow depth of field, photorealistic, editorial quality

“Single-source directional light from the left” creates the raking light that makes texture pop — this is a real photography technique and the AI generator responds to it accurately. “Deep shadows revealing texture” reinforces that instruction. For any product where material quality is a selling point, this type of prompt does more work than three lifestyle shots combined.

5. Multiple product variants — consistency across colors — critical for e-commerce where you need the same shot in six colorways.

Minimalist water bottle, product photography, isolated on light grey gradient background, centered composition, studio lighting with soft fill, shot from front elevation, photorealistic commercial product photo, 4K — color variant: forest green

Minimalist water bottle, product photography, isolated on light grey gradient background, centered composition, studio lighting with soft fill, shot from front elevation, photorealistic commercial product photo, 4K — color variant: matte white

Keep every element of the prompt identical except the color. Same background, same lighting description, same angle, same shot type. This is how you maintain subject consistency across a product line — the prompt structure is your style guide. If you’re using AI Studio or the API, you can template this and loop through a list of color values programmatically.

Pro tip ✅

When generating product variants, always note the exact prompt you used for the first approved image and treat it as locked. Changing even one word — “soft lighting” vs “softbox lighting” — can shift the overall look enough to make variants feel like they came from different photoshoots.

6. Packaging shot with text — for products where the box or label is part of the brand experience.

Premium chocolate bar packaging, dark navy blue box with gold foil typography reading "MAISON DULCE", product photography, isolated on black velvet surface, single overhead spotlight creating a dramatic pool of light, luxury brand aesthetic, photorealistic, high contrast

Gemini’s text rendering handles short phrases on packaging with reasonable accuracy — keep the text under 20 characters for best results, avoid complex scripts, and always verify the output carefully before using it. “MAISON DULCE” in caps is easier to render cleanly than lowercase cursive. The velvet surface + spotlight combination is a classic luxury product photography setup that reads as premium without being fussy.

Warning ⚠️

Text rendering in AI image generators is still imperfect. Always zoom in and read every character in any generated shot that includes text. Subtle letter transpositions and font inconsistencies happen. For legally important text — ingredients, certifications, barcodes — never rely on AI-generated images. Generate the shot, then composite the real text in post.

7. Editorial-style product shot — for brands that want to look like they belong in a magazine rather than a marketplace.

Perfume bottle, editorial product photography, placed on the edge of a marble bathroom shelf, late afternoon sunlight casting long geometric shadows, out-of-focus plants in the background, film grain, muted color palette, Vogue-style luxury aesthetic, photorealistic

“Film grain” and “muted color palette” shift the output away from the hyper-clean AI look toward something that reads as analog and considered. “Vogue-style luxury aesthetic” is a surprisingly effective instruction — the generator has processed enough of that reference material that the phrase carries real meaning. The geometric shadows from late afternoon sun add visual interest without props doing the heavy lifting.

Pro tip ✅

If your product shot comes back looking too clean — the AI equivalent of over-processing — add “slight lens imperfection,” “subtle vignette,” or “natural color grading” to the prompt. Controlled imperfection reads as authenticity.

The Iteration Workflow

Generating one image and using it is rarely the move. The real workflow looks like this: generate three to five variants of the same prompt by running it multiple times, pick the best structural composition, then refine with a second prompt that addresses specific problems — too much background clutter, wrong shadow direction, surface reflection needs adjusting. Use Nano Banana’s image editing capabilities to make targeted changes rather than regenerating from scratch every time you want to tweak the foreground shadow.

In AI Studio, you can also experiment with temperature settings — lower temperature produces more predictable, conservative interpretations of your prompt, which is usually what you want for product photography. Higher temperature gets creative in ways that are exciting for lifestyle shots but problematic when a client needs six consistent catalog images by Thursday.

Pro tip ✅

Generate at the highest available resolution setting. Downsampling a sharp 4K image for web thumbnails always looks better than upscaling a small image. Start big, size down — never the reverse.

Pro tip ✅

For product lines with up to five hero items, Nano Banana’s subject consistency features can keep the same character or object treatment coherent across a series. Establish the visual language in your first approved image, then explicitly reference it: “same lighting and background treatment as reference” when continuing a set.

What This Actually Saves You

A product photography session with a professional photographer, studio rental, and basic retouching runs anywhere from a few hundred dollars for a simple setup to several thousand for a full day with multiple setups. Nano Banana doesn’t fully replace a studio for complex products, but for early-stage brands, small catalogs, social media content, and A/B testing different visual directions before committing to a shoot, it closes the gap significantly. The prompts in this guide are a starting point, not a ceiling — every product category has its own visual conventions, and the more specifically you can describe those conventions in your prompt, the closer your output gets to something a real photographer would bill you for.

The tool is available now. Your product catalog isn’t going to photograph itself.

promptyze

See Full Bio