GPT Image 2.5 output format: png, jpeg or webp on Sume

GPT Image 2.5 on Sume returns png, jpeg or webp through output_format. Which to pick for transparency and speed, and what output_compression does on Sume today.

3 min readSume
All posts

GPT Image 2.5 on Sume can return png, jpeg or webp, set with output_format; png is the safe choice for transparency and jpeg or webp for smaller files. OpenAI's guide lists PNG as its default and recommends JPEG over PNG for time-sensitive uses. Sume's output_compression (0 to 100) is not served yet, so a request that sets it fails.

From Sume's Image API docs and OpenAI's image generation guide, read 2026-09-29.

Which format should I choose?

Read 2026-09-29. From Sume's docs and OpenAI's guide.
FormatAlpha channelUse when
pngYes, with background: "transparent"Cutouts, graphics, lossless
webpYes, with background: "transparent"Web delivery, smaller files
jpegNoPhotos where size or speed matters

How do I set it?

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 bowl of ramen, overhead shot",
    "output_format": "webp",
    "quality": "medium"
  }'

Can I set output_compression?

Not yet. The docs say output_compression and seed are in the schema but no model advertises them, so sending one returns 400 unsupported_parameter. Re-encode in your own pipeline if you need a specific compression level.

Sources

Related posts

More in Developers

All Developers posts

Written by Sume