Fast text-to-image
Short enough per run that trying five directions costs less than deliberating over one.
OpenAI fast image model
Turn ideas into images fast with GPT Image 2.5 Flare, or build new versions from a reference — made for social assets and quick creative exploration.

Switch between text-to-image and reference editing. Credits update before you run anything.
This page runs at medium quality with 1K, 2K and 4K output and a choice of aspect ratios, returning one PNG per run. Text-to-image costs 23, 39 or 69 credits by resolution; image editing costs 33, 49 or 79.
Text-to-image starts from a description. Image editing needs one reference upload.
A sentence is enough to start. Run it, look at what came back, then adjust one thing at a time.
Choose the aspect ratio and resolution; the credit cost updates as you do. The result stays in your history.
GPT Image 2.5 Flare is the variant to reach for while you are still deciding what the image should be.
Short enough per run that trying five directions costs less than deliberating over one.
Headlines and short copy come through legibly, which most fast models still get wrong.
Upload one image and spin off alternates that keep the composition but change the treatment.

Feed, story and banner from the same prompt, without recropping by hand.
Draft at 1K while exploring, then run the winner larger.
The price of the run is on screen before you press generate — 23 to 69 credits for text-to-image.
Real output from this model at the settings this page uses.




Posts and stories that need a fresh visual this afternoon, not next week.
The same item in several settings, so the campaign has options to choose between.
Putting a rough idea on screen quickly enough that the team can react to it.
OpenAI
23 Click to start
OpenAI
40 Click to start
Describe a direction and let GPT Image 2.5 Flare put it on screen while it is still worth reacting to.
Start creating