gpt-image-1 shuts down Oct 23, 2026: what to send on Sume

OpenAI's deprecations page lists gpt-image-1 with a shutdown date of October 23, 2026 and recommends gpt-image-2.5-sunburst or gpt-image-2.5-flare. On Sume, send openai/gpt-image-2.5 (Flare) or openai/gpt-image-2.5-sunburst to POST /v1/images, or use model: "sume/auto".

Dates and replacements are from OpenAI's page (read 2026-09-30); Sume ids are from the Images docs. The same page lists gpt-image-1-mini and gpt-image-1.5 for December 1, 2026.

Which dates and replacements does OpenAI list?

What model id do I send to Sume?

Sume ids use an org/slug form: openai/gpt-image-2.5 is Flare and openai/gpt-image-2.5-sunburst is Sunburst. Both support text-to-image, up to 16 image references, an optional mask_url, and background set to auto, transparent or opaque. ChatGPT Image 2 remains selectable.

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 matte ceramic mug on a linen cloth" }'

What if I pass an id Sume does not know?

The API returns 404 model_not_found with the message "Unknown image model id." and points to /v1/images/models. Do not assume the retired OpenAI id still maps to something; list the catalog and pick from it. Bare legacy ids such as gpt-image-2 and nano-banana-2 are accepted as aliases for their org/slug equivalents.

Can I let Sume choose?

Yes: model: "sume/auto" lets Image Router pick the family. It is not listed in GET /v1/images/models, and job.model stays sume/auto. Pin an id instead when you need repeatable style. For the difference between the two ids, see Flare vs Sunburst.

What should I do before October 23?

Change the model string, run a few of your real prompts, and compare the outputs by eye. Check any parameters your old requests used against the catalog, because unsupported ones return 400 unsupported_parameter.