GPT Image 2.5 Flare vs Sunburst: what each model page says

GPT-Image-2.5 Flare and GPT-Image-2.5 Sunburst are two image models OpenAI lists side by side. OpenAI's Flare page calls it "our fastest model for high-quality, everyday image generation"; its Sunburst page calls it "our most capable model for image generation and editing". Both take text and image input, both return images, and both use the same token prices, so the choice is about the job, not the price sheet.

On Sume you pick between them with the model field of POST /v1/images: openai/gpt-image-2.5 is Flare and openai/gpt-image-2.5-sunburst is Sunburst. Model facts below come from OpenAI's two model pages and Sume's Image API docs, read 2026-09-29.

What does OpenAI say about each model?

Only what its model pages say. Neither page, as read, gives a side-by-side quality ranking, so this post makes no claim about which output looks better. Test both on your own prompts.

What does Sume list for each id?

The same request surface. The Image API docs say both ids support text-to-image, up to 16 image references, an optional mask_url, and background: auto|transparent|opaque, with quality from auto to max and high as the default when you omit it. They share the same token rates, so a request costs the same on either id at the same size and quality.

The docs add that Sume's Auto model routing continues to use Flare. Sunburst is only used when you name it.

How do I switch between them in one request?

Change only the model string. Everything else in the body stays valid, which makes a fair test cheap: send the same prompt and references to both ids and compare the results yourself.

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-sunburst",
    "prompt": "Keep the person and pose. Change the jacket to dark green.",
    "input_references": [
      { "type": "image_url", "image_url": { "url": "https://example.com/portrait.jpg" } }
    ],
    "quality": "medium",
    "aspect_ratio": "auto"
  }'

Which one should I start with?

If you mostly generate from text, start with Flare: it is the id Sume's Auto routing uses, and OpenAI positions it for everyday generation. If your workflow is a chain of edits to the same picture, add Sunburst to the test, because OpenAI describes it as its most capable model for generation and editing. That is a starting point drawn from vendor descriptions, not a measured result.