Choose Sunburst for precision or Flare for speed. Both generate and edit at up to 4K with as many as 16 references.
| ASTRIA |
Model launch / 09.26 |
|
|
Two OpenAI image models / ready to try
GPT Image 2.5,
two ways to work.
Choose Sunburst when exact edits and visual fidelity matter. Choose Flare when you want a faster generation loop. Both generate and edit from text and references inside Astria.
|
|
|
01 / Precision
Sunburst handles the careful revisions.
Start with Sunburst for complex image work, controlled edits, and stronger visual fidelity. Generate from text, edit with references or a mask, and combine up to 16 reference images when the brief has many parts to reconcile.
Try GPT Image 2.5 Sunburst in Generate →
|
|
|
02 / Speed
Flare keeps the iteration loop moving.
Reach for Flare when you want to explore directions and produce variations faster. It uses the same Astria controls for text generation, reference editing, masks, resolution, aspect ratio, and quality.
Try GPT Image 2.5 Flare in Generate →
|
|
Pick by job
Precision or speed?
Both models support 1K, 2K, and 4K output plus five quality levels from Low through Max. Choose the model that fits the next pass, then add your prompt and references.
|
|
The model links open Generate with Sunburst or Flare already selected. Quality choices are Low, Medium, High, Extra High, and Max.
|
|
Astria