Ideal inputs
Use one product or object photo with the entire subject visible; describe the desired shadow, crop, and amount of white space.
Place an uploaded subject into a simple white studio-style scene while keeping the subject and its silhouette recognizable.
Use one product or object photo with the entire subject visible; describe the desired shadow, crop, and amount of white space.
Choose a clear source and mention whether the framing should be centered, close, or leave room for copy.
Specify a pure white background, shadow softness, camera height, and any parts that must not change.
Check the silhouette, contact shadow, highlights, and label detail against the white background.
GPT Image 2 generates a white-background composition guided by the source and the studio direction.
The workflow is an image edit, not a pixel-perfect background replacement. Reflections, transparent materials, and exact product geometry may vary.
No. A clearly separated subject on another background can be used, but the model may reinterpret difficult edges.
The prompt asks it to remain stable, but generated edits can alter geometry or labels. Verify the result against the original.