What is GPT Image 2.5 on vid2vid?

GPT Image 2.5 is an AI image generation and editing model available in vid2vid. Choose Flare for prompt-led exploration or Sunburst for a refinement pass, and use either with reference editing.

What is the difference between Flare and Sunburst?

Flare is the default speed-oriented option for exploring visual directions. Sunburst is the refinement-oriented option to try when a direction is already clear and you want a more considered pass.

Can I use GPT Image 2.5 for image editing?

Yes. Open image-to-image, upload up to 16 JPEG, PNG, or WebP reference images (20MB per image), and explain what should change or stay consistent.

Which image sizes and backgrounds are supported?

The generator offers 1K, 2K, and 4K output choices plus transparent, opaque, or automatic backgrounds. The 27:16, 16:27, 9:8, and 8:9 aspect ratios are available only at 1K.

Should I start with text-to-image or image-to-image?

Start with text-to-image when the visual is still an idea. Use image-to-image when a product, person, composition, or style reference needs to guide the edit.

Create your first 5-second AI video free with a source clip up to 5 seconds.Try for Free

vid2vid

GPT Image 2.5

GPT Image 2.5 AI Image Generator

Create from prompts or references with fast Flare drafts and Sunburst refinement. Hero image: illustrative reference inspiration.

Flare speed modeSunburst refinement modeText to imageReference editing1K · 2K · 4KTransparent · opaque · auto backgrounds

Create with GPT Image 2.5 Flare Edit a reference image

Illustrative image of an orange toy robot in warm golden-hour light

GPT Image 2.5 workspace

Create or edit with GPT Image 2.5

Flare is selected by default for text-to-image. Switch to image-to-image when a reference should guide the edit, then choose the GPT Image 2.5 variant that fits the stage of your work.

Ready to create with this model

Open the generator when you are ready to upload a reference or write a prompt.

Open generator

Two GPT Image 2.5 modes, one clear workflow

Choose a mode by intent, then keep the same prompt and reference workflow while you explore ideas, refine details, and prepare assets for a product, social, or editorial brief.

Flare for fast exploration

GPT Image 2.5 Flare is the default selection for quickly exploring composition, subject, lighting, and art direction from a prompt or reference.

Sunburst for premium refinement

Switch to GPT Image 2.5 Sunburst when the direction is clear and you want a more considered refinement pass for the same creative brief.

Text-to-image when the brief is still open

Start with words when you are defining a new product scene, social concept, campaign frame, or editorial image from scratch.

Image-to-image when the reference matters

Upload a reference and describe the change when composition, subject identity, product details, or an existing visual language should guide the result.

What GPT Image 2.5 supports

Choose the model, input mode, output size, and background treatment that fit the image brief.

Flare is selected by default for text-to-image, and Sunburst is available when the creative direction is established.

Choose text-to-image or image-to-image, with 1K, 2K, and 4K output options and transparent, opaque, or automatic backgrounds.

Upload up to 16 JPEG, PNG, or WebP reference images, with a 20MB limit per image.

Where GPT Image 2.5 fits best

Use the mode that matches the stage of the visual brief, from first direction through a shareable campaign asset.

Product launch visuals

Explore product hero compositions, packaging scenes, launch concepts, and alternate art directions with a prompt or a controlled reference.

Social content systems

Build a set of social-first visual directions with consistent subjects, clear crops, background choices, and sizes that match the publishing plan.

Editorial and campaign art direction

Develop mood, lighting, wardrobe, location, and composition references before a shoot, layout, video treatment, or wider creative review.

How to use GPT Image 2.5 on vid2vid

Keep the input strategy simple, make the output intent explicit, and refine only when the direction is worth carrying forward.

Choose text-to-image or image-to-image

Use text-to-image for a fresh visual direction. Use image-to-image when a reference should anchor the subject, composition, or style change.

Pick Flare or Sunburst

Keep Flare selected while you are exploring. Move to Sunburst when you want to refine a direction that is already working.

Set output intent and generate

Choose 1K, 2K, or 4K where available, set a transparent, opaque, or automatic background, then review the result in your workspace history.

GPT Image 2.5 FAQ

Answers about Flare and Sunburst, input modes, output choices, references, and the image workflow.

Turn a visual brief into a usable direction

Open the GPT Image 2.5 workspace with Flare selected, or jump into image-to-image when your reference is ready.

Open GPT Image 2.5 Flare Try Sunburst refinement