Reference-photo fidelity
Transforms people, places, and other subjects into new settings, visual styles, and compositions while better preserving recognizable features, natural lighting, and richer textures.
ChatGPT Images 2.5
Freemium
ChatGPT Images 2.5 is an image-generation and editing model for turning ideas, sketches, prompts, and reference photos into more personalized images. It supports iterative creative work in ChatGPT and production image workflows through the API.
What is ChatGPT Images 2.5?
ChatGPT Images 2.5 is OpenAI’s image-generation and editing model for creating images from text prompts, sketches, and reference photos. It is designed for creative workflows that require recognizable subjects, natural lighting and textures, precise edits, and consistency across multiple rounds of changes.
The model can transform reference photos into new settings, visual styles, and compositions while preserving distinctive subject features. It can also make focused changes to an image—such as updating a product, background, or piece of copy—while preserving the surrounding composition and brand treatment. OpenAI says generation latency is reduced by up to 50% compared with Images 2.0.
In ChatGPT, Sketch lets users draw directly as a reference for a final image. Templates provide starting points for formats including flyers and product photos, while comments placed directly on images support more focused editing. Users can also share prompts so others can try the same ideas with their own photos and details.
ChatGPT Images 2.5 is available in ChatGPT, ChatGPT Work, and Codex across desktop, mobile, and web. For developers, OpenAI offers GPT-Image-2.5 Flare in the API with the same quality, editing, and speed improvements, and GPT-Image-2.5 Sunburst for more precise detailed creative work with longer generation times.
What can ChatGPT Images 2.5 do?
Precision editing
Targets the requested element—such as a product, background, or piece of copy—while keeping other image details, composition, and brand treatment more stable.
Multi-turn consistency
Carries earlier edits through longer conversations so successive changes build on existing work without degrading image quality as readily.
Complex visual instruction handling
Interprets more detailed visual instructions and supports coherent layouts, real-world information, and transparent backgrounds.
Sketch, templates, and image comments
ChatGPT workflows include drawing a sketch as a visual reference, starting from templates for formats such as flyers and product photos, and placing comments directly on images for focused revisions.
API model options
Developers can use GPT-Image-2.5 Flare for the model’s quality, editing, and speed improvements, or GPT-Image-2.5 Sunburst for more precise detailed creative work with longer generation times.
Use Cases
“Reference-led portraits and composites”
Use a person’s reference photo to create a portrait in a different setting or style, or combine familiar subjects into a new composition while retaining recognizable details.
“Product and brand asset revisions”
Update a product, background, or text element in an existing image without rebuilding the entire asset, helping preserve the surrounding composition and brand treatment.
“Flyers and product-photo concepts”
Start from a template when creating common visual formats, then refine the result with prompts, image comments, or additional edits.
“Sketch-to-image ideation”
Draw a rough idea directly in ChatGPT and use the sketch as a reference for developing a more polished final image.
“Programmatic creative workflows”
Integrate image generation and targeted revisions into an application through the API, using Flare for the standard Images 2.5 improvements or Sunburst when detailed precision is more important than generation time.
Frequently Asked Questions
Where can ChatGPT Images 2.5 be used?
It is available to ChatGPT, ChatGPT Work, and Codex users across desktop, mobile, and web. Developers can access related GPT-Image-2.5 models through the API.
Can it edit an existing image over several turns?
Yes. The model is designed to follow focused editing instructions across multiple turns, preserve earlier changes more reliably, and build on an existing image without degrading quality as readily.
What are Flare and Sunburst?
GPT-Image-2.5 Flare brings the Images 2.5 improvements in quality, editing, and speed to the API. GPT-Image-2.5 Sunburst is intended for more precise detailed creative work and has longer generation times.
Can I use a sketch or reference photo as input?
Yes. ChatGPT’s Sketch feature lets you draw directly as a reference, and the model can use reference photos to create new settings, styles, and compositions while preserving more of the original subject’s distinctive features.
Quick Facts
- Category
- AI image generation and editing
- ChatGPT access
- ChatGPT, ChatGPT Work, and Codex
- Supported surfaces
- Desktop, mobile, and web
- API models
- GPT-Image-2.5 Flare and GPT-Image-2.5 Sunburst
- Generation speed
- Up to 50% lower latency than Images 2.0, according to OpenAI
- Input types
- Text prompts, sketches, and reference photos
ChatGPT Images 2.5 Alternatives
Imejis.io
www.imejis.io
Imejis.io is an image generation platform with a remote MCP server and REST API for creating, editing, and rendering structured designs. It helps AI-agent users, developers, and automation workflows turn templates or structured data into PNG, JPEG, WebP, and single-page PDF outputs.
豆绘AI
douhuiai.com
豆绘AI is an online AI image creation and editing platform for e-commerce and design, supporting white-background, scene, interior renderings, and common post-processing.
Upsampler
upsampler.com
Upsampler is an online AI platform for generating, editing, and upscaling images and videos in one workspace. Pay per credit, with full API access and a commercial use license on every plan.
Creativio AI
creativio.io
AI product photography that turns basic images into polished marketing visuals
简单AI
ai.sohu.com
简单AI is Sohu’s online AI toolkit for image generation, writing, and basic photo editing.
Recraft
www.recraft.ai
Recraft is an AI image and design platform for creatives and teams, with raster/vector generation, custom styles, mockups, editing, and API workflows.