Original historical site image
AI Models & Assistants / Knowledge library
GPT-Image-2.5 Sunburst
Our most capable model for image generation and editing
Why learn GPT-Image-2.5 Sunburst?
Offers the documented OpenAI model capabilities through hosted.
The problem it solves
Provides model inference for supported text / image inputs.
Where it came from
GPT-Image-2.5 Sunburst: Our most capable model for image generation and editing Published aliases: gpt-image-2.5-sunburst, gpt-image-2.5-sunburst-2026-09-08. Input: text, image. Output: image. Checked 2026-10-13. Model capabilities are distinct from tools available in a particular agent.
Build your understanding
Work through these exercises in order. Try the task before opening a reference.
- Explain what GPT-Image-2.5 Sunburst does.
- Identify a use case for GPT-Image-2.5 Sunburst.
- Describe a limitation of GPT-Image-2.5 Sunburst.
- Name a compatible tool for GPT-Image-2.5 Sunburst.
- Compare GPT-Image-2.5 Sunburst against an alternative.
Documentation and sources
- Primary source / model card (developers.openai.com)
Catalog research notes
- See the AI Workbench for source dates, pricing evidence and configuration details.
Connect the ideas
Built on
Real World Example
Architectural Visualization Refinement
A boutique architectural firm specializing in historical building renovations.
The business problem
The firm struggles to integrate modern structural additions into existing historical site photographs, often resulting in mismatched lighting, perspective errors, or loss of architectural detail during manual editing.
How this topic solves it
The firm utilizes GPT-Image-2.5 Sunburst to perform precise inpainting, allowing them to mask specific areas of historical site photos and generate photorealistic structural additions that inherit the lighting and texture of the original environment.
- Upload the original high-resolution historical site photograph to the Image API.
- Define the target region for the new structure using an alpha-channel mask.
- Submit the masked image and a descriptive prompt to the gpt-image-2.5-sunburst model.
- Select the 'xhigh' quality setting to ensure fine architectural details are preserved.
- Review the generated output and iterate using the same model if structural adjustments are needed.
The flow shows how an image is processed through the model to produce a final architectural render.
Defining the area for the new structure
Model processing with xhigh quality
Integrated architectural design
- Source Photo → apply mask → Masking
- Masking → input masked image → GPT-Image-2.5 Sunburst
- GPT-Image-2.5 Sunburst → generate output → Final Render
Expected business outcome
The firm achieves seamless visual integration of new designs into legacy environments, significantly reducing the time spent on manual post-production blending.
Limitations and checks
The model may struggle with extreme perspective shifts if the input image lacks sufficient depth information; verify this by checking the alignment of generated structural lines against the original vanishing points.