Animate Photo AI
首页历史记录
替换角色Wan 2.2 AnimateWan Replace Character让老照片动起来照片转视频Talking Photo AI
升级解锁更快生成速度和更多积分

GPT Image 2.5 Editor

登录

模型

参考图片

0/16

Common edit directions

点击填入

画面比例

分辨率

background

0 可用 credits

编辑结果

上传原图，并描述一个明确的修改。

生成完成后，编辑后的图片会显示在这里。

上传原图，并描述一个明确的修改。

Home/Models/GPT Image 2.5 editor

Prompt-based image editing

GPT Image 2.5 Editor

The GPT Image 2.5 editor lets you upload an image, describe a precise change, and compare a generated result before you download or animate it. Choose GPT Image 2.5 Flare for everyday variations or Sunburst for more explicit creative direction.

Source upload, prompt, model, and output controls in this GPT 2.5 editor
Upload a source, write one focused edit, choose a GPT Image 2.5 image-to-image model, and review the generated result.

Continue your workflow

AI photo editor workbench

AI Photo Editor

Edit photos with a broader model selection.

Old photo restoration example

Old Photo Restoration

Repair faded scans before you animate them.

Photo to video result

Photo to Video

Turn a checked image into a short motion clip.

Editor capabilities

What you can do with the GPT Image 2.5 editor

Precise object edits

Add, remove, or replace a defined object while asking the model to preserve the surrounding composition.

Background and scene changes

Move a portrait or product into a cleaner setting with explicit lighting, perspective, and subject-preservation instructions.

Reference-guided variations

Add reference images for identity, style, or visual direction when one source image is not enough.

High-resolution review

Compare 1K, 2K, and 4K output options, then inspect faces, hands, small text, and edges before export.

Four-step workflow

How to use the GPT Image 2.5 editor

  1. 1

    Upload a source image

    Use a clear JPG, PNG, or WebP image up to 12 MB.

  2. 2

    Describe one focused edit

    Say what should change and what must remain unchanged.

  3. 3

    Choose Flare or Sunburst

    Use Flare for everyday variations or Sunburst for more explicit creative direction.

  4. 4

    Review and download

    Check faces, hands, text, edges, and small details before downloading or animating.

Model and output choices

Choose the right GPT Image 2.5 image editor model

ModelBest forInputOutputCredits
GPT Image 2.5 FlareEveryday edits and quick creative variations.1 source + up to 15 reference images (16 URLs total).1K, 2K, or 4K image output.1K: 30; 2K: 45; 4K: 70 credits
GPT Image 2.5 SunburstEdits that need more explicit creative direction.1 source + up to 15 reference images (16 URLs total).1K, 2K, or 4K image output.1K: 30; 2K: 45; 4K: 70 credits

Provider: Kie image-to-image endpoints. Animate Photo AI is an independent interface, not the official OpenAI product. View the Kie model reference and pricing page. Last checked September 11, 2026.

Typical use cases

When to use GPT Image 2.5 image editing

Product photo cleanup

Upload a product image, remove clutter, and create a cleaner e-commerce scene without rebuilding the entire composition.

Portrait and identity-preserving edits

Use a portrait plus references to adjust the setting, styling, or lighting while explicitly protecting the face, pose, and clothing.

Creative campaign variations

Create several controlled background, color, or layout variations, then select the strongest frame for social, ads, or motion.

Quality checks

Review generative edits before you publish

GPT Image 2.5 editing follows instructions but does not guarantee pixel-level preservation. Use one major change per pass and keep the original source file.

  • Faces and identity

    Compare the result with the source before publishing.

  • Hands and thin edges

    Inspect fingers, jewelry, and narrow object boundaries.

  • Logos and small text

    Check brand marks and text for altered characters.

  • Reference images

    Use references for guidance; exact copying is not guaranteed.

Questions

GPT Image 2.5 editor FAQ

What is the GPT Image 2.5 editor?

GPT Image 2.5 Editor is an independent Animate Photo AI image-to-image workflow. Upload one source, describe one focused change, optionally add references, compare the result, and download the generated file.

What is the difference between Flare and Sunburst?

Flare is positioned for everyday edits and quick variations. Sunburst is positioned for edits that need more explicit creative direction. These are product guidance labels, not a published latency or quality benchmark.

Can I edit an image with reference images?

Yes. The workbench accepts a source image and optional reference images for identity, style, or scene guidance. The Kie image-to-image endpoints support up to 16 input image URLs in one request.

How much does a GPT Image 2.5 edit cost?

Each GPT Image 2.5 Flare or Sunburst edit costs 30 credits at 1K, 45 credits at 2K, or 70 credits at 4K on this workbench. Review the visible quote before generation because account plans determine credit value.

Is this the official OpenAI image editor?

No. This is an independent Animate Photo AI workflow that connects to GPT Image 2.5 image-to-image models through Kie. OpenAI and GPT Image are referenced only to identify the model family.

Try the GPT Image 2.5 editor

Upload a source image, describe one clear change, and compare Flare with Sunburst in the workbench.

Edit an image

Last updated September 11, 2026