Best Buy product shot guide: 1500x1500 and what to ask Sume

What image size does Best Buy ask for?

Best Buy's Digital Assets Product Shot Guide-Generic asks for a minimum of 1500 by 1500 pixels at 300 DPI, cropped close to the product on a white background, and cropped images must still meet the minimum size. The default image is preferably a 45 degree or front-facing shot, named with the product's UPC.

A caveat on scope: this is the generic shot guide on Best Buy's partners site. It says it covers assets for e-commerce, print, digital media and in-store advertising, and its contact line is a Best Buy Canada studio address. I did not read a page specific to Best Buy Marketplace sellers, so treat this as Best Buy's general digital asset spec and confirm the marketplace rule during onboarding.

The guide's numbers

The guide separates a minimum shot set from an enhanced one. Both start with the same first five views; the enhanced set adds close-up, size reference, lifestyle and a 360 degree video.

Which Sume settings meet 1500 by 1500?

On ChatGPT Image 2.5 (openai/gpt-image-2.5), 1500 is not a legal custom edge because both edges must be multiples of 16 and 1500 divided by 16 is 93.75. The next multiple of 16 above it is 1600, and 1600 by 1600 is 2,560,000 pixels, inside the 655,360 to 8,294,400 range and under the 3840 maximum edge. So a 1600 by 1600 custom image_size clears Best Buy's 1500 minimum with room to crop.

Nano Banana 2 (google/nano-banana-2) works by tier instead: resolution 0.5K, 1K, 2K or 4K at aspect ratio 1:1. The Sume docs do not state exact pixel counts for those tiers on this page, so measure the output before trusting it against a 1500 pixel floor.

What to ask Sume for, and what it cannot do

Ask for the main 45 degree view and the front view as separate calls from the same reference photo, with an opaque background and a PNG or JPEG. Sume's output formats are png, jpeg, webp and svg, so JPEG, the third format on Best Buy's list, is available; EPS, TIFF and PSD are not. The 300 DPI value is metadata in the file rather than a request field on the Image API, so set it in your own tooling after generation.

``json { "model": "openai/gpt-image-2.5", "prompt": "Same product, dynamic 45 degree angle, cropped close, plain white background, no text, no shadow props", "input_references": [{"type": "image_url", "image_url": {"url": "https://example.com/front.jpg"}}], "image_size": "1600x1600", "quality": "high", "background": "opaque", "output_format": "jpeg" } ``

Naming by UPC and the rest of the shot list are your pipeline's job. For the cost side of batches, see the per-SKU pipeline cost post.