GPT Image 2.5 transparent background: how to request a cutout by API

To get a transparent background from GPT Image 2.5 on Sume, send background: "transparent" and an output_format of png or webp. Sume lists background (auto, opaque, transparent) on the two 2.5 ids only, and OpenAI's guide says transparent backgrounds require png or webp; JPEG has no alpha channel.

Read from Sume's Image API docs and OpenAI's image generation guide on 2026-09-29.

What does the request look like?

Describe the subject alone, on its own, and let the model cut it out. The response's media_type should be image/png or image/webp.

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": "a single ceramic mug, studio lighting, isolated subject",
    "background": "transparent",
    "output_format": "png",
    "quality": "high"
  }'

What are the background values?

Check the edges before you ship

Look at the result on a dark and a light backdrop. Hair, glass and soft shadows are where a generated cutout can leave halos. For an existing photo, a dedicated background remover may be a better tool: see remove image background API.

What happens with ChatGPT Image 2?

background is not listed on openai/gpt-image-2, so it returns 400 unsupported_parameter. The docs point to Image 1.0 with transparency: true for transparent stills on other routes.