Generate from a text prompt, at 1K or 2K, change the scene. Keep the subject.
| Founded year: | 2026 |
| Country: | China |
| Funding rounds: | Not set |
| Total funding amount: | Not set |
Description
image25.photo is a workspace for making and editingpictures with OpenAI's GPT Image 2.5 models. Write a sentence and get an image,
or hand it one you already have and say what should change — the setting, the
style, the composition — and the subject still comes back recognizable. That
consistency is the headline of the 2.5 release: OpenAI reports more of the
original detail surviving an edit, along with better infographic accuracy and
layout — the difference between a chart you can read and one that is only
decorative.
Two models sit behind the same brief. **Flare** is the faster default, the one to
reach for while an idea is still moving. **Sunburst** takes longer and holds
detail, for a layout you have already decided on. Both bill at the same rate, so
switching between them costs nothing but time.
