Kaufland product image requirements and what to ask Sume for

Kaufland Global Marketplace's image rules (1024 px, 2048 px zoom, 5 to 10 pictures, white first photo) and the Sume Image API request that fits them.

5 min readSume
All posts

What are Kaufland's product image requirements?

Kaufland Global Marketplace asks for at least 1024 px in height or width, at least 2048 px for the zoom function, at most 4096, files up to 10 MB, and 5 to 10 pictures per product. The first photo must show the whole product with a clear silhouette on a white background.

The prohibited list is long: collages, shadows, frames or white space, text, logos, symbols or flags, delivery or price information, promotional lettering such as 'Sale', and official seals such as Stiftung Warentest.

Kaufland Global Marketplace, Product Data Guidelines (read 2026-10-03).
RuleWhat the page says
ResolutionMinimum 1024 px in height or width; 2048 px for zoom; maximum 4096
File sizeMaximum 10 MB
Formatspng, jpg, jpeg, webp, tiff, gif
Quantity5 to 10 pictures from different perspectives or detail shots
First photoWhole product, clear silhouette, white background, not a snapshot
ContentOnly products and accessories in the scope of delivery
ProhibitedCollages, shadows, frames, white space, text, logos, symbols, flags, price or delivery information, promotional content, official seals

What do you ask Sume for?

Target the zoom size, 2048x2048. On openai/gpt-image-2.5 that is legal custom pixels: both edges are multiples of 16 (2048 is 128 x 16), the largest edge is under the 3840 cap, and 4,194,304 pixels is inside the documented 655,360 to 8,294,400 range. Kaufland's 4096 maximum is above what that model accepts, so you will never hit it.

Ask for a clean cut-out look in the prompt, and keep the reference product unchanged. Both png and jpeg are in Sume's output_format list and in Kaufland's format list.

curl -X POST https://api.sume.com/v1/images \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "openai/gpt-image-2.5",
  "prompt": "Same product, unchanged, whole product visible, white background, no shadow, no frame, no text, no logo",
  "input_references": [{"type": "image_url", "image_url": {"url": "https://example.com/packshot.png"}}],
  "image_size": "2048x2048",
  "quality": "high",
  "background": "opaque",
  "output_format": "png",
  "n": 1
}'

How do you fill a five-to-ten picture gallery?

Kaufland wants different perspectives and detail shots, and only the first photo carries the white-background rule. Per the catalog code, most image models take n as 1 to 4, so plan two or three calls, each with its own angle in the prompt, rather than one call for ten.

Keep accessories to what ships in the box; Kaufland says only products and accessories in the scope of delivery may appear.

Kaufland rules (read 2026-10-03) against Sume request fields from the Image API docs.
Kaufland ruleSume fieldValue
2048 px zoomimage_size2048x2048
White first photobackground plus promptopaque, 'white background'
5 to 10 picturesn per call1 to 4, several calls
No collagePromptOne product per frame
Accessories in delivery onlyinput_referencesOnly photos of what ships

What to verify before upload?

Open each file and look for shadows, borders and stray text; the prompt asks, it does not guarantee. Check pixel size and file size against the table, and keep the first photo as the plain white one. Images that mix the rules, such as a lifestyle scene with a 'Sale' badge, are exactly what the page lists as prohibited.

Sources

Related posts

More in Use cases

All Use cases posts

Written by Sume