gpt-image-2-5 model example image

GPT Image 2.5 in DaVinci — subjects that stay recognizable

Create with GPT Image 2.5 in DaVinci, a creative workspace that brings together image models from different AI providers. Work from reference photos: move a person, a product, or a place into a new setting or style while the details that identify it carry through.

Why creators choose GPT Image 2.5 in DaVinci

Start Creating

Subjects stay themselves in DaVinci

Faces, products, and places keep the features that make them recognizable when you move them into a different setting or composition with GPT Image 2.5 in DaVinci.

Edits that stay put in DaVinci

Change one part of the frame in DaVinci and the rest is left alone. Add a mask when you want the edit confined to an area you marked yourself.

Up to 16 references in DaVinci

Combine as many as sixteen reference images in one edit through DaVinci, so the subject, the product, and the setting can each come from a different photo.

Built for real layouts in DaVinci

Generate up to 3840 px on the long edge in square, landscape, or portrait from DaVinci, so the result drops into a layout without being upscaled first.

How to create with GPT Image 2.5 in DaVinci

Generate images, edit them with reference photos, and refine your prompt with GPT Image 2.5 directly inside DaVinci AI.

Start Creating

Who is GPT Image 2.5 for in DaVinci?

GPT Image 2.5 is built for marketers, ecommerce teams, and creators, and it is ready to use in DaVinci.

Try GPT Image 2.5
DaVinci for designers and creatives

DaVinci for designers and creatives

Put the same character or object into new scenes and styles in DaVinci without redrawing it from scratch every time the concept moves.

DaVinci for brands and marketing teams

DaVinci for brands and marketing teams

Keep the product looking like the product across a whole campaign, then adjust single elements with a mask in DaVinci when feedback comes in.

DaVinci for content creators and businesses

DaVinci for content creators and businesses

Generate thumbnails, banners, and ecommerce imagery in DaVinci from photos you already have, then edit them as the brief changes.

Try GPT Image 2.5

Describe your idea, add a reference photo if you have one, and generate it with GPT Image 2.5 in DaVinci.

Start Creating
GPT Image 2.5 example image

Powerful features in DaVinci

Here’s why creators choose GPT Image 2.5 in DaVinci.

Start Creating

Reference fidelity in DaVinci

Subjects carry their identifying detail into new settings, styles, and compositions when you edit with GPT Image 2.5 in DaVinci.

Reference-based editing in DaVinci

Bring in as many as 16 reference photos in one edit through DaVinci and describe the change you want made.

Masked editing in DaVinci

Mark an area with a mask in DaVinci and the change is confined to it, so the rest of the frame is untouched.

High-resolution output in DaVinci

Up to 3840 px on the long edge, in square, landscape, or portrait, up to a 3:1 ratio, generated in DaVinci.

Frequently Asked Questions

Yes. Reuse the same reference image or saved character/object for GPT Image 2.5 in DaVinci. State the fixed details—product shape, logo, and label—and change one scene element at a time. Compare each output with the source because small details can drift.

Yes. Use the source image with a GPT Image 2.5 variant, then write one specific edit such as “change only the background to a bright office; keep the person’s face and pose.” Choose GPT Image 2.5 Flare or Sunburst in DaVinci, then describe one precise edit and the details that must remain unchanged.

In DaVinci’s Image workspace, choose Add References for GPT Image 2.5. You can add a reference image, character, or object and explain its role in the prompt. For multiple sources, state which image supplies the subject and which supplies the setting; compare the result with each source.

Inspect the output at full size. Check text, faces, hands, edges, colors, and composition, then revise the prompt or edit the result where needed.

DaVinci provides GPT Image 2.5 through the Flare and Sunburst variants. Choose the variant that fits the image workflow, then compare both with the same prompt when you need to evaluate the difference.

The credit cost can differ between the model variants and can change with image settings. Select each variant with the same image count and resolution, compare the totals shown before generation, and weigh the result against the cost.

If a GPT Image 2.5 edit changes more than intended, identify the exact region to change and describe what must remain untouched. Keep the source image fixed and compare the next version with it.

The current picker shows GPT Image 2.5 Flare and Sunburst as separate variants, while GPT Image 2 is listed separately. Compare the same prompt across the versions you can select, focusing on text, subject consistency, detail, and cost.

GPT Image 2.5 is available in DaVinci through the Flare and Sunburst variants. Each generation uses credits, and the exact cost for your selected settings is shown before you generate.

Use the same prompt in GPT Image 2.5 and one alternative. Compare composition, detail, text rendering, consistency, input controls, and credit cost before choosing.

Design with DaVinci

Create with momentum. Bring your vision to life.

Sign Up