Back to all models

OpenAI / GPT Image

GPT Image 2.6 Flare

Fast-iteration generation and reference editing with 1K–4K output.

GPT Image 2.6 Flare is the faster-iteration member of the 2.6 family. This platform generates one PNG image or edits one uploaded reference, with fifteen aspect ratios at every resolution tier. Choose low, medium or high quality; the displayed credit quote is fixed before submission.

What you can do

  • Text-to-image and single-reference editing
  • 1K, 2K and 4K with fifteen aspect ratios at every tier
  • Low, medium and high quality; one PNG result per model task
  • Compare with Sunburst using the same prompt in All-in mode
Graphic poster featuring a mechanical hummingbird — reference artwork, not an attributed model output.
Reference imageVisual inspiration only. This image is not presented as an output from the selected model.

Ideas to try

  • Explore poster layouts and marketing directions
  • Iterate on product scenes or reference-based visual changes

Practical tips

  • Start at 1K and medium quality; review the quote before raising tiers.
  • Describe what to preserve when editing. This platform currently delivers static PNG images, not GIF or SVG.

Supported controls on AIO

These specifications come from AIO’s operation profiles. Controls can differ by tool and resolution; final credit cost is shown in Studio before you run.

Generate

Available
Aspect ratios
1:1 / 16:9 / 9:16 / 4:3 / 3:4 / 3:2 / 2:3 / 5:4 / 4:5 / 21:9 / 2:1 / 1:2 / 3:1 / 1:3 / 9:21
Resolution
1K / 2K / 4K
Quality
Low / Medium / High
Task count options
1
Images per task
1
AIO prompt presets
Cinematic / 3D clay / Cyber neon

Style presets add curated words to your prompt; they are not native model styles.

Edit

Available
Aspect ratios
1:1 / 16:9 / 9:16 / 4:3 / 3:4 / 3:2 / 2:3 / 5:4 / 4:5 / 21:9 / 2:1 / 1:2 / 3:1 / 1:3 / 9:21
Resolution
1K / 2K / 4K
Quality
Low / Medium / High
Task count options
1
Images per task
1
AIO prompt presets
Cinematic / 3D clay / Cyber neon

Style presets add curated words to your prompt; they are not native model styles.

Requires your own input image in Studio. Reference art on this page is not uploaded for you.