Walmart's AI content rule: accurate, truthful, rights-compliant
Walmart's image guide says AI-generated content must be accurate, truthful and rights-compliant, with no logos or non-English text. Review each output.

Walmart's image guide says AI-generated content must be accurate, truthful, rights-compliant, and not misleading. Main images cannot carry watermarks, your name or logo, or Walmart or other retailer logos, and should not include languages other than English. So the prompt is the easy part; the review of each image is what keeps a listing live.
This is from Walmart's image guidelines, last updated May 12, 2026 and read on 2026-10-02. Size and format rules are covered in the related post.
What does the page say about content?
The content rules are separate from the size rules.
| Item | What the page says |
|---|---|
| AI-generated content | Must be accurate, truthful, rights-compliant and not misleading |
| Main image background | Seamless white (255/255/255 RGB) |
| Main image marks | No watermarks, your name or your logo |
| Retailer logos | No Walmart or other retailer or platform logos or names |
| Language | No languages other than English |
| Last updated | May 12, 2026 |
What is a safe review checklist?
Run each generated image through the same questions before upload.
- Does the product match the real one in shape, colour, parts and size?
- Did the model add a logo, label or foreign-language text?
- Is the background plain white with no stray shadow shapes?
- Do you have the rights to every element in the picture?
A prompt that keeps the product fixed
Send the real product photo as input_references to the Image API and say what must not change. Then compare the result against the reference by eye.
import os, requests
r = requests.post(
"https://api.sume.com/v1/images",
headers={"Authorization": f"Bearer {os.environ['SUME_API_KEY']}"},
json={
"model": "sume/auto",
"prompt": "Place this exact blender on a seamless pure white background, soft shadow, no text, no logos, do not change the product",
"input_references": [
{"type": "image_url", "image_url": {"url": "https://example.com/products/blender.jpg"}}
],
"aspect_ratio": "1:1",
},
timeout=60,
)
body = r.json()
if r.status_code == 200:
print(body["data"][0]["url"])
else: # 202: poll the job
print(body["data"]["status_url"])
Sources
Related posts
More in Use cases
- Walmart Media Library video: 100 MB limit and 200-character title
Walmart's Media Library takes videos of 100 MB or less with a title up to 200 characters, and restricts alcohol, tobacco and firearms. Check a Sume clip first.
- Walmart minimum four images per listing: a four-shot batch per SKU
Walmart's image guide asks for at least four images per listing. A four-shot plan per SKU on Sume: main on white, angle, detail, in-use, plus the checks to run.
- Walmart Sponsored Videos run 5 to 45 seconds: cut yours to length
Walmart Connect says Sponsored Videos run between five and 45 seconds. How to trim a finished product video to that window with Sume's video-trim endpoint.
- Wan 3.0 30-second drafts at 480p, then the 1080p final: cost on Sume
Alibaba lists Wan 3.0 at $0.05, $0.10 and $0.20 a second. A 30-second shot is $1.50, $3.00 or $6.00 at list; Sume's list x 1.25 makes it $1.875, $3.75 or $7.50.
Written by Sume