Wayfair image guide: five shot types, 1000x1000 minimum

Wayfair's Imagery 101 guide says images need to be at least 1,000 x 1,000 pixels to be featured, and 2,000 x 2,000 will show every detail. It names five shot types: single-product silhouette, zoomed-in environmental, zoomed-out environmental, functional and dimensional. You can plan the whole set as five prompts against one clean packshot with Sume's Image API, then check each frame against what the product really looks like.

Everything about Wayfair below comes from its Imagery 101 page, read on 2026-10-02. That page does not state file formats, video rules or AI rules, so check Wayfair's current supplier requirements for those before upload.

What are the five Wayfair image types?

The guide asks for variety: different shot types and angles, with size and features visible. Here is the set and a plain-language shot brief for each.

Which of these can AI make, and which need a real photo?

The silhouette is your real packshot; do not regenerate it, because it is the ground truth the rest are built from. Environmental scenes are the natural fit for a reference-guided edit: the packshot goes in as input_references and the prompt describes the room. Functional and dimensional shots carry claims about how the product works and how big it is, so verify them against the real item and its spec sheet.

Sume's Image API takes up to 10 public HTTPS reference images, an aspect_ratio, a resolution tier of 512, 1K, 2K or 4K, and n for several variants. Wayfair's 1,000 x 1,000 floor is met by a square 1K output; ask for 2K if you want the 2,000 x 2,000 the guide recommends. A model only accepts the tiers its catalog lists, so read GET /v1/images/models first.

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": "Zoomed-out room scene: this armchair by a window in a bright living room, true color and proportions, no extra furniture styling that changes the product",
        "input_references": [
            {"type": "image_url", "image_url": {"url": "https://example.com/packshots/armchair.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"])

A QA pass before you submit

Generated furniture drifts: leg count, seam lines, cushion thickness and color are the usual failures. Put the packshot and each scene side by side and reject any frame where the product itself changed.

• Count legs, drawers, handles and cushions against the real item.
• Compare color to the packshot under neutral light.
• Reject any scene that implies a size or function the product does not have.
• Export at 1,000 x 1,000 or higher and keep the file for your own records.