AI Product Photo Editing for Marketplaces

Remove backgrounds, enhance, expand and upscale product photos for Amazon, Shopify, Coupang and Naver listings without changing the product. Steps, credits, checks.

AI-generated green steel bottle on a white studio sweep beside a laptop showing the cut-out product

The fastest way to turn ordinary product photos into marketplace-ready images is a fixed four-step edit: enhance the light and color, remove the background, expand the canvas to the ratio each channel wants, then upscale. In Cut Marketing’s Photo Editor Studio that pipeline costs 43 credits per photo (35 without Enhance) and runs on single images or batches. The rule that matters most: edits may change the background, framing and resolution, never the product itself.

In short:

  • Edit in this order: enhance → remove background → expand → upscale.
  • Background removal, Enhance and Expand are 8 credits per image each, and 4K upscale is 19.
  • Amazon’s main image needs a pure white background and a product that fills most of the frame; for Coupang, Naver Smart Store and Shopify, check each seller’s current image guide.
  • Keep the original file and compare every edit against it, especially edges, labels and transparent materials.
  • Generate new scenes only when editing cannot give you the image you need.

What each AI edit actually does

Photo Editor Studio groups four operations. Each one is listed in the model catalog with its credit cost:

Edit What it does Credits Output
Background remover Cuts the product or subject out of its background 8 per image Source ratio
Enhance Corrects light, color and clarity 8 per image Source ratio
Expand Outpaints the canvas to a new ratio 8 per image 1:1, 4:5, 9:16 or 16:9
4K upscale Re-renders the image at 4K with sharper detail 19 per image Source ratio

You can run an edit on one photo or on a batch. Uploads are limited to 10 MiB per file. The studio shows the exact cost before you submit, and a failed job releases its reserved credits automatically.

The order of operations, and why it matters

1. Enhance first

Fix exposure, white balance and clarity on the full original photo. The cut-out tool works better on a well-lit image with clear contrast between product and background. If you correct color after the cut-out, you may shift the product against a white background you have already set.

2. Remove the background

Run background removal on the enhanced image. This gives you the product alone, ready to place on white or on a new scene. Inspect the edge: hair-thin details, cables, straps, transparent glass and soft shadows are where cut-outs most often fail.

3. Expand to the target ratio

Marketplaces and ad placements want different shapes. Expand outpaints the canvas to 1:1, 4:5, 9:16 or 16:9 so the product is not cropped to fit. Square is the usual starting point for listing images; 4:5 and 9:16 are useful when the same product photo also feeds social ads.

4. Upscale last

Run 4K upscale once the framing is final. Upscaling earlier means you pay for, and inspect, pixels you might later cut away. Upscaling also estimates detail, so doing it last keeps that estimation to a single, reviewable step.

Marketplace notes

Rules differ by marketplace and change over time. Treat this table as a planning aid and confirm the details in each seller center before you upload.

Marketplace Main image notes
Amazon Pure white background (RGB 255,255,255); product fills about 85% of the frame; at least 1000 px on the longest side so zoom works
Coupang Check the current image guide in the seller center
Naver Smart Store Check the current image guide in the seller center
Shopify Your theme decides the display; keep one consistent ratio across the catalog

Secondary images usually allow more freedom: scale cues, in-use scenes, detail close-ups and comparison layouts. For Korean marketplaces in particular, the detail page carries much of the selling work; our guide to Coupang and Naver Smart Store listing images covers the section structure.

Keep the product truthful

AI edits are fast, which makes it easy to approve something that no longer matches what the customer receives. Check every edited image for:

  • Color. Enhance should correct light, not change the product’s actual color. Compare against the physical product or a calibrated reference.
  • Labels and text. Upscaling can sharpen letters into shapes that look like text but are not. Read every label.
  • Shape and size cues. Expand should add background, not new product parts. Check that the outpainted area does not add props that imply a bundle.
  • Edges. Look for halos, missing straps, clipped caps and transparent areas filled in with grey.
  • Shadows. A floating cut-out on white can look unnatural; a fake shadow that changes the apparent size is misleading.

Keep the original file for every SKU. If a buyer, marketplace or your own team questions an image, you can show what the photo looked like before editing.

When to generate instead of edit

Editing is the right tool when the original photo is accurate and you only need a cleaner background, a different ratio or more resolution. It is the wrong tool when you need an image that does not exist yet: a lifestyle scene, a model holding the product, a seasonal background or a detail-page banner.

For those, use generation and then review against the real product:

  • Listing Images creates product image drafts from a reference photo for marketplace review.
  • Reference Images generates product images guided by a reference image and a written brief.
  • Image Generator creates static ads, packshots and backgrounds from text, from 2 credits ($0.02) for a FLUX.3 Klein image.
  • AI Detail Page Generator builds the long detail page in the widths the Korean marketplaces use: 780 px for Coupang and 860 px for Naver SmartStore.

Image models are priced per output. Nano Banana 2.1, for example, costs 6 credits at 1k and 8 at 2k. A generated image is a new image, not a retouched photo, so it needs the same product-accuracy review as any other AI output. The AI product image generator page compares the generation options.

Worked credit examples

All numbers below use the per-image costs shown in the catalog. Your exact quote appears before each run.

Job Steps per photo Credits per photo Photos Total
Clean main images Background remover + Expand + 4K upscale 8 + 8 + 19 = 35 40 1,400
Same, with light correction Enhance + Background remover + Expand + 4K upscale 8 + 8 + 8 + 19 = 43 40 1,720
Cut-outs only Background remover 8 40 320
Trial run Background remover 8 3 24

New accounts receive 30 one-time trial credits: enough to cut out three photos (24 credits), or to run one cut-out plus Expand (16 credits), and judge the quality on your own products. The full 35-credit pipeline needs more than the trial balance. Without a plan, batches are limited to 50 images and video to 720p; paid plans start at $9 a month and raise both. See pricing for the current details.

Step-by-step batch workflow

  1. Collect originals. One folder per product, highest-quality files under 10 MiB each.
  2. Test on three photos. Run the full pipeline on a small sample with different materials, for example matte, glossy and transparent.
  3. Review the sample. Check color, labels, edges and shadows against the originals.
  4. Run the batch. Use the same settings for the whole catalog so images look consistent.
  5. Spot-check. Open every image with fine detail or text at full size.
  6. Export and name. Use SKU-based file names and keep originals alongside edits.
  7. Upload and verify. Preview each listing on desktop and mobile before publishing.

Pre-upload checklist

  • The product color matches the physical product.
  • Labels, logos and text are readable and unchanged.
  • No halo, missing part or grey fill around the edges.
  • Expand did not add props or product parts.
  • The main image follows the marketplace’s current background and framing rules.
  • Resolution is high enough for zoom where the marketplace supports it.
  • Originals are saved next to the edited files.

FAQ

In what order should I edit product photos with AI?

A reliable order is enhance, remove the background, expand to the target ratio, then upscale. Correct light first, cut out on the corrected image, set the canvas, and enlarge last so you do not upscale pixels you later discard.

Can AI background removal produce an Amazon main image?

It can produce a clean cut-out, but you still need to place it on pure white (RGB 255,255,255), make the product fill about 85% of the frame and check edges. Always confirm against Amazon’s current image requirements.

Does AI upscaling add real detail?

No. Upscaling estimates new pixels, which can invent texture, sharpen text oddly or soften fine edges. Compare the result with the source before publishing.

How many credits does a typical marketplace photo cost?

In Photo Editor Studio, background removal, Expand and Enhance are 8 credits each and 4K upscale is 19. Background removal, Expand and 4K upscale together cost 35 credits per photo, and the exact cost is shown before you run it.

When should I generate a new scene instead of editing a photo?

Edit when you already have an accurate photo and only need a cleaner background, ratio or resolution. Generate a new scene for lifestyle or detail-page images, then review it against the real product.

Try it on your own catalog

Open Photo Editor Studio with Google sign-in and run your first three cut-outs on the trial credits. Read the Photo Editor Studio tool page for supported edits, or check pricing before you plan a full catalog.

Keep reading

AI-generated tabletop studio with a glass bottle, a white sneaker and a fuzzy knit beanie on a pale grey sweep beside a phone on a mini tripod
Product photography

Remove Background From Product Photos: Free and Batch Methods

Remove a product photo background with a free built-in tool for a few images, or batch AI for a catalog. Methods compared, hard cases fixed, Amazon rules.

AI-generated tabletop photo set with an amber dropper bottle on a stone plinth, a softbox, a sketchbook of thumbnail compositions and three colour swatches
Product photography

AI Product Photography Prompts: 15 Templates

A six-part formula and 15 copy-paste AI product photography prompts for packshots, lifestyle scenes, macros and ad layouts, with the fix lines we use against logos and bad hands.

AI-generated grid of nine square listing photos of an amber dropper bottle on a grey table
Marketplaces

Coupang and Naver Smart Store Images with AI

How to prepare Coupang and Naver Smart Store main images, extra images and detail pages with AI: cutouts, page sections, and what must stay a real photo.

Make the next ad in minutes.

Start with a product image or a short brief. The studio shows the credit cost before you generate.

Open studio