Build a campaign visual
Describe the focal subject and reserve space for your copy. Use the result as the image layer of an editable layout.
OpenAI Model family
GPT Image models, from first idea to final edit.
Explore OpenAI’s image model family. Compare GPT Image generations and variants, trace their releases, and choose a starting point for image creation and editing.
3 supported modelsRelease history & guidesReviewed 2026-10-05
The collection
GPT Image models
All GPT Image options currently supported in Quickmake. Release dates refer to provider announcements or launch notes; unverified dates are marked below.
3 of 3 models
GPT Image 2.5
2 models available
| Model | Workflow | Released | Credits | Generate |
|---|---|---|---|---|
GPT Image 2.5 Sunburst OpenAI’s quality-oriented 2.5 variant for evaluating detailed visual briefs. Quality focused | Text to imageImage editing | Announcement | 11/ image | |
GPT Image 2.5 Flare OpenAI’s speed-oriented 2.5 variant for exploring and refining directions. Speed focused | Text to imageImage editing | Announcement | 11/ image |
GPT Image 2
1 model available
| Model | Workflow | Released | Credits | Generate |
|---|---|---|---|---|
GPT Image 2 An earlier generation to revisit when comparing an established image workflow. Earlier generation | Text to imageImage editing | Announcement | 43/ image |
Credits shown are for one 1024 × 1024 image with default settings and no references. Settings can change the quote; review it before generating. Quickmake credits and plans ↗
The bigger picture
A family for creating and refining images
GPT Image brings image generation and editing into a prompt-led workflow. For a campaign, that can mean creating the initial scene, changing its setting, and bringing the selected image into a layout with editable copy. Choosing a model is one part of that process; defining what must stay consistent is just as useful.
The generation number identifies a release, while names such as Sunburst and Flare identify choices within it. Start with the table to compare supported versions. The release guides go deeper into individual options, so you can revisit an established workflow without treating every GPT Image model as interchangeable.
Ways to use GPT Image
Refine a product image
Attach your reference and ask for a specific change, such as a new setting. Inspect the product silhouette and label before accepting it.
Compare Sunburst and Flare
Use the same brief, dimensions, and quality setting. Choose based on the results and credit estimates for your actual task.
Where to start
- Choose between two GPT Image 2.5 variants without changing your design workspace.
- Try text-to-image generation and reference-based image editing from the same model family.
- Use the linked review for an editorial comparison and a practical evaluation brief.
What to check
- Provider quality and speed positioning does not predict the winner for every prompt.
- Generated lettering and preserved reference details still need visual review.
- Image generation produces a raster asset; build editable text and layout elements on the canvas.
GPT Image questions
It includes every GPT Image image model currently supported in Quickmake. The provider may offer other models, tools, or modes that are not integrated here.
Dates refer to the linked provider announcement or launch notes, not the date a model was added to Quickmake. If a dated release source is not confirmed, the table says so and links to documentation instead.
OpenAI positions Sunburst around quality and Flare around speed. Quickmake supports both. Compare them on the same brief rather than assuming one is best for every image.
GPT Image is the image model family used by the integration. Generating in Quickmake uses your Quickmake account and credits; it does not use your ChatGPT subscription.
The guide linked below covers Sunburst versus Flare, practical prompts, editing checks, pricing considerations, and alternatives. It reviews documented capabilities rather than claiming a scored hands-on benchmark.
You can read this page and explore the prompts without signing in. Generating an image requires a Quickmake account and sufficient AI credits. The generator shows the estimate before you submit.
Explore related models and tools
Sources and availability
Reviewed 2026-10-05. The options above reflect Quickmake’s supported integration; provider documentation may describe additional features.